DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
This action is responsive to the amendments and remarks received 03 June 2026. Claims 1, 3 - 8 and 10 are currently pending.
Claim Objections
Claim 10 is objected to because of the following informalities: Line 4, lines 6 - 7 and lines 7 - 8 of claim 10 each recite, in part, “the training set” which appears to contain inconsistent claim terminology and/or a minor informality. The Examiner suggests amending line 4, lines 6 - 7 and lines 7 - 8 of claim 10 to --the infrared ship detection training set-- in order to maintain consistency with line 4 of claim 1 and to improve the clarity and precision of the claim. Appropriate correction is required.
The objections to claims 1, 2, 5 and 8 - 10, due to minor informalities, are hereby withdrawn in view of the amendments and remarks received 03 June 2026.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1, 3 - 8 and 10 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 1 recites the limitation "the divided data set in Visual Object Classes (VOC) format" (emphasis added) in lines 5 - 6. There is insufficient antecedent basis for this limitation in the claim.
Claim 1 recites the limitation "the trained infrared ship target detection model" (emphasis added) in line 14. There is insufficient antecedent basis for this limitation in the claim. The Examiner suggests amending line 13 of claim 1 to --obtain [[an]] a trained infrared ship target detection model--.
Claim 6 recites the limitation "the input feature layers having different resolutions" (emphasis added) in lines 5 - 6. There is insufficient antecedent basis for this limitation in the claim.
Claim 7 recites the limitation “the i-th layer” in line 8. There is insufficient antecedent basis for this limitation in the claim.
Claim 10 is rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention because it is unclear as to which infrared ship target detection model “the infrared ship target detection model” recited on lines 2 - 3, along with subsequent recitations of “the infrared ship target detection model”, are referencing. Are they referring to the “infrared ship target detection model” recited on lines 10 - 11 of claim 1 or the “infrared ship target detection model” recited on line 13 of claim 1? Clarification and appropriate correction are required. For purposes of examination, the Examiner will treat model “the infrared ship target detection model” recited on lines 2 - 3 of claim 10, along with subsequent recitations of “the infrared ship target detection model”, as referencing the “infrared ship target detection model” recited on lines 10 - 11 of claim 1.
Claim 10 recites the limitation “the infrared ship target detection model trained” (emphasis added) in lines 10 - 11. There is insufficient antecedent basis for this limitation in the claim.
Claims 3 - 5 and 8 are also rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, due to being dependent upon a rejected base claim(s) but would be withdrawn from the rejection if their base claim(s) overcome the rejection.
Response to Arguments
Applicant’s arguments with respect to claim(s) 1, 3 - 8 and 1 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Applicant's arguments filed 03 June 2026 have been fully considered but they are not persuasive.
On page 11 of the remarks the Applicant’s Representative argues that claim 1 is not rendered obvious by Wang et al. in view of Ye et al. The Applicant’s Representative argues that the “Examiner has not sufficiently shown why a person of ordinary skill in the art would have been motivated to specifically modify Wang with Ye so as to obtain the claimed improved YOLOv7-based infrared ship target detection method having the recited combination of a MobileNetv3 network, a bidirectional weighted feature pyramid network, an attention mechanism, and an optimized loss function.” Furthermore, the Applicant’s Representative argues that “the Examiner's given rationale is generic and hindsight approach and does not adequately explain the specific claimed combination and does not provide any rationale as to how a skilled person would be motivated to arrive at the claimed method in view of Wang with Ye.” Therefore, the Applicant’s Representative argues that claim 1 is not rendered obvious by Wang et al. in view of Ye et al.
The Examiner respectfully disagrees.
Initially, the Examiner asserts that instant claim 1 is currently rejected under 35 U.S.C. 103 as being unpatentable over Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al.
Furthermore, the Examiner asserts that Wang et al., particularly their target detection model, was modified to utilize an infrared maritime ship data set for training in order to obtain an infrared ship target detection model useable to detect a maritime ship as taught by Ye et al. The Examiner asserts that Wang et al. was not modified by Ye et al. to obtain a target detection model “having the recited combination of a MobileNetv3 network, a bidirectional weighted feature pyramid network, an attention mechanism, and an optimized loss function.”
Furthermore, in response to applicant’s argument that there is no teaching, suggestion, or motivation to combine the references, the examiner recognizes that obviousness may be established by combining or modifying the teachings of the prior art to produce the claimed invention where there is some teaching, suggestion, or motivation to do so found either in the references themselves or in the knowledge generally available to one of ordinary skill in the art. See In re Fine, 837 F.2d 1071, 5 USPQ2d 1596 (Fed. Cir. 1988), In re Jones, 958 F.2d 347, 21 USPQ2d 1941 (Fed. Cir. 1992), and KSR International Co. v. Teleflex, Inc., 550 U.S. 398, 82 USPQ2d 1385 (2007). In this case, the Examiner asserts that at least the knowledge generally available to one of ordinary skill in the art provides sufficient motivation to modify Wang et al. with Ye et al. The Examiner asserts that one of ordinary skill in the art would recognize that target detection models can be trained to detect any type of target of interest and modifying Wang et al. with Ye et al. as proposed would enhance the base device of Wang et al. by allowing for it to be employed in a wider variety of situations and/or utilized in an increased number and variety of related and applicable applications and/or environments, such as for the detection of target vehicles, for example ships, in infrared images, thereby improving its overall appeal, usefulness and marketability to potential end-users.
In addition, in response to applicant's argument that the examiner's conclusion of obviousness is based upon improper hindsight reasoning, it must be recognized that any judgment on obviousness is in a sense necessarily a reconstruction based upon hindsight reasoning. But so long as it takes into account only knowledge which was within the level of ordinary skill at the time the claimed invention was made, and does not include knowledge gleaned only from the applicant's disclosure, such a reconstruction is proper. See In re McLaughlin, 443 F.2d 1392, 170 USPQ 209 (CCPA 1971).
Therefore, the Examiner asserts that a person of ordinary skill in the art would have been sufficiently motivated to modify Wang et al. with Ye et al. to obtain an infrared ship target detection model useable to detect a maritime ship and that instant claim 1 is rendered obvious by Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al.
On page 11 of the remarks the Applicant’s Representative argues that claims 2 - 6 “are not rendered obvious by the cited references” at least because claims 2 - 6 “recite additional specific limitations directed to data processing, network reform, MobileNetv3 structure, bidirectional weighted feature pyramid network configuration, and weighted feature fusion operations, which are not taught or suggested in the particular claimed combination relied upon by the Office Action.”
The Examiner respectfully disagrees.
The Examiner asserts that Applicant's arguments fail to comply with 37 CFR 1.111(b) because they amount to a general allegation that the claims define a patentable invention without specifically pointing out how the language of the claims patentably distinguishes them from the references.
Furthermore, the Examiner asserts that instant claim 2 has been cancelled, that instant claims 3 - 6 are currently rejected under 35 U.S.C. 103 as being unpatentable over Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. and that the additional specific limitations of instant claims 3 - 6 are disclosed by the currently cited references, see at least section 18 of the instant Office Action included herein below. Therefore, the Examiner asserts that instant claims 3 - 6 are rendered obvious by the currently cited references.
On page 11 of the remarks the Applicant’s Representative argues that “the Examiner has not sufficiently established that one of ordinary skill in the art would have specifically incorporated the weighted feature fusion formulas relied upon from Tan into the combined Wang and Ye system in the particular claimed manner.” Therefore, the Applicant’s Representative argues that “claim 7 is not rendered obvious by Wang in view of Ye and Tan.”
The Examiner respectfully disagrees.
Initially, the Examiner asserts that instant claim 7 is currently rejected under 35 U.S.C. 103 as being unpatentable over Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. further in view of Tan et al.
Furthermore, the Examiner asserts that they have sufficiently established that one of ordinary skill in the art would have incorporated the weighted feature fusion formulas from Tan et al. into the combined base device of Wang et al. in view of Ye et al., currently Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. The Examiner asserts that the “Supreme Court in KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398, 415-421, 82 USPQ2d 1385, 1395-97 (2007) identified a number of rationales to support a conclusion of obviousness which are consistent with the proper ‘functional approach’ to the determination of obviousness as laid down in Graham.” The Examiner asserts that the Supreme Court recognized “Simple substitution of one known element for another to obtain predictable results” as a rationale that may support a conclusion of obviousness, see MPEP § 2143. The Examiner asserts that one of ordinary skill in the art would have incorporated the weighted feature fusion formulas from Tan et al. into the combined base device of Wang et al. in view of Ye et al., currently Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al., at least to simply substitute the undisclosed weighted feature fusion formulas of Wang et al. for the weighted feature fusion formulas of Tan et al. in order to fuse the feature information of the feature layers of the combined base device. In addition, the Examiner directs the Applicant’s attention to MPEP § 2143, section 26 of the Office Action mailed 30 March 2026, and section 19 of the instant Office Action included herein below.
Therefore, the Examiner asserts that one of ordinary skill in the art would have incorporated the weighted feature fusion formulas from Tan et al. into the combined base device of Wang et al. in view of Ye et al., currently Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al.
On page 12 of the remarks the Applicant’s Representative argues that “the Examiner has not sufficiently established that one of ordinary skill in the art would have specifically modified the combined Wang and Ye system to employ the recited SENet structure with a soft attention mechanism” as claimed. Therefore, the Applicant’s Representative argues that “claim 8 is not rendered obvious by Wang in view of Ye and Xiangrong.”
The Examiner respectfully disagrees.
Initially, the Examiner asserts that instant claim 8 is currently rejected under 35 U.S.C. 103 as being unpatentable over Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. further in view of Xiangrong et al.
Furthermore, the Examiner asserts that they have sufficiently established that one of ordinary skill in the art would have modified the combined base device of Wang et al. in view of Ye et al., currently Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al., to employ the SENet channel attention mechanism of Xiangrong et al. The Examiner asserts that the “Supreme Court in KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398, 415-421, 82 USPQ2d 1385, 1395-97 (2007) identified a number of rationales to support a conclusion of obviousness which are consistent with the proper ‘functional approach’ to the determination of obviousness as laid down in Graham.” The Examiner asserts that the Supreme Court recognized “Simple substitution of one known element for another to obtain predictable results” as a rationale that may support a conclusion of obviousness, see MPEP § 2143. The Examiner asserts that one of ordinary skill in the art would have modified the combined base device of Wang et al. in view of Ye et al., currently Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al., to employ the SENet channel attention mechanism of Xiangrong et al. at least to simply substitute the SE attention mechanism of Wang et al. for the SENet channel attention mechanism of Xiangrong et al. in order to enable the combined base device to focus on and strengthen the representational power of useful information and features. In addition, the Examiner directs the Applicant’s attention to MPEP § 2143, section 27 of the Office Action mailed 30 March 2026, and section 20 of the instant Office Action included herein below.
Therefore, the Examiner asserts that one of ordinary skill in the art would have modified the combined base device of Wang et al. in view of Ye et al., currently Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al., to employ the SENet channel attention mechanism of Xiangrong et al.
On page 12 of the remarks the Applicant’s Representative argues that the previously cited references “do not teach or suggest the specifically recited training and verification process as a whole, including cross-verifying an average accuracy change and loss change trend based on the training set and the verification set and adjusting the training process until the recited changes tend to be stable, in the particular claimed manner.” Therefore, the Applicant’s Representative argues that “claim 10 is not rendered obvious by Wang in view of Ye and Zheng.”
The Examiner respectfully disagrees.
The Examiner asserts that Applicant's arguments fail to comply with 37 CFR 1.111(b) because they amount to a general allegation that the claims define a patentable invention without specifically pointing out how the language of the claims patentably distinguishes them from the references.
Furthermore, the Examiner asserts that instant claim 10 is currently rejected under 35 U.S.C. 103 as being unpatentable over Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. further in view of Zheng et al.
Moreover, the Examiner asserts that, at least, Zheng et al. disclose cross-verifying an average accuracy change and loss change trend based on a training set and the verification set and adjusting the training process until the recited changes tend to be stable, see at least section 21 of the instant Office Action included herein below and/or section page 12 section 4.2 - page 16 line 5 and figures 11 - 13 of Zheng et al. wherein they disclose that they “randomly divide the dataset into training set, validation set and test set according to the ratio of 6:2:2. Considering the device memory setting and usage, the training batch size is set to 8, the initial learning rate is 0.001, and the Adam algorithm is used to optimize the loss function, and the number of iterations is set to 300”, that to “objectively evaluate the performance of the algorithms, this paper uses precision, recall, mean average precision (MAP), and F1 to evaluate the performance of different models” and that the “model was trained and validated using a dataset of six types of ships commonly found in inland waterways collected. The loss function plays an important role in the training process as it reflects the relationship between the true and predicted values. The smaller the loss, the closer the prediction is to the true value and the better the performance of the model. The loss function for the MC-YOLOv5s model training process was calculated and plotted as shown in Fig 11. It can be observed that after 300 iterations, the loss function of the model keeps decreasing and reaches convergence, resulting in a better training model.”
The Examiner asserts that, as shown herein above and in the cited sections and figure, Zheng et al. disclose cross-verifying an average accuracy change and loss change trend based on the training set and the verification set and adjusting the training process until the recited changes tend to be stable at least because they disclose utilizing the Adam algorithm, which adaptively adjusts the learning rate during the training process, to optimize the loss function, because they train and validate their model on a training set and a validation set, respectively, until the loss function of their model reaches convergence, becomes stable, and because figure 11 of Zheng et al. illustrate that the average accuracy change and loss change trend of their model during training on the training set and the validation set becoming stable. Therefore, the Examiner asserts that instant claim 10 is rendered obvious by the currently cited references.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Claims 1 and 3 - 6 are rejected under 35 U.S.C. 103 as being unpatentable over Wang et al., “An improved YOLOv7 method for vehicle detection in traffic scenes”, IEEE, 35th Chinese Control and Decision Conference (CCDC), May 2023, pages 766 - 771 in view of Jing Ye, Zhaoyu Yuan, Cheng Qian, and Xiaoqiong Li, “CAA-YOLO: Combined-Attention-Augmented YOLO for Infrared Ocean Ships Detection”, Sensors, Vol. 22, Issue 10, May 2022, pages 1 - 23, herein referred to as “Ye et al.”, in view of Chen et al. U.S. Publication No. 2023/0188671 A1 in view of Zhen-wei Wu, Ming-hao Liu, Cheng-xiu Sun and Xin-fa Wang, “A dataset of tomato fruits images for object detection in the complex lighting environment of plant factories”, Data in Brief, Vol. 48, Article 109291, June 2023, pages 1 - 6, herein referred to as “Wu et al.”.
- With regards to claim 1, Wang et al. disclose a method for detecting a target based on an improved YOLOv7 (You Only Look Once version 7), (Wang et al., Pg. 766 Abstract, Pg. 766 § I ¶ 3, Pg. 767 § II ¶ 1, Pg. 768 Subsection C ¶ 1, Pg. 769 § III - Subsection B, Pg. 770 Subsection D - Section “Discussion and Analysis”, Pg. 770 Fig. 7) comprising the following steps: obtaining a data set; (Wang et al., Pg. 766 § I ¶ 3, Pg. 769 § III - Pg. 770 Section “Discussion and Analysis”) dividing the data set into a detection training set, a verification set and a test set according to a ratio of 7:2:1, (Wang et al., Pg. 769 § III - Pg. 770 Section “Discussion and Analysis” [“the data set is divided into training set, validation set and test set according to the ratio of 7:2:1”]) and performing data enhancement preprocessing on the data set, (Wang et al., Pg. 766 Abstract, Pg. 769 § III - Subsection B [“the mosaic data enhancement is used,”]) wherein the data enhancement preprocessing includes Mosaic image enhancement; (Wang et al., Pg. 766 Abstract, Pg. 769 § III - Subsection B [“the mosaic data enhancement is used,”]) reforming a YOLOv7 network structure based on a MobileNetv3 network and a bidirectional weighted feature pyramid network, (Wang et al., Pg. 766 Abstract, Pg. 766 § I ¶ 3, Pg. 767 § II - Pg. 768 Subsection B, Pg. 770 Section “Discussion and Analysis”) and obtaining a target detection model by introducing an attention mechanism and an optimized loss function; (Wang et al., Pg. 766 Abstract, Pg. 766 § I ¶ 3, Pg. 767 § II - Pg. 768 Subsection C, Pg. 770 Section “Discussion and Analysis”) training and verifying the target detection model based on the data set to obtain a target detection model trained; (Wang et al., Pg. 766 Abstract, Pg. 766 § I ¶ 3, Pg. 769 § III - Pg. 770 Section “Discussion and Analysis”) and detecting a target based on the trained target detection model. (Wang et al., Pg. 766 Abstract, Pg. 766 § I ¶ 3, Pg. 767 § II ¶ 1, Pg. 769 § III - Pg. 770 Section “Discussion and Analysis”, Pg. 770 Fig. 7) Wang et al. fail to disclose explicitly detecting an infrared ship target; an infrared maritime ship data set; converting the divided data set in Visual Object Classes (VOC) format into a data set in YOLO format, wherein the data enhancement preprocessing includes left-right flipping and image zooming, an infrared ship target detection model; and detecting a maritime ship. Pertaining to analogous art, Ye et al. disclose a method for detecting an infrared ship target based on an improved YOLO (You Only Look Once), (Ye et al., Pg. 1 Abstract, Pg. 2 First-Full Paragraph - Second-Full Paragraph, Pg. 5 § 3 - Pg. 7 § 3.3 ¶ 1, Pg. 11 § 4 - Pg. 13 § 4.2.2, Pg. 13 Fig. 9, Pg. 15 Figs. 10 & 11, Pg. 16 Fig. 12, Pg. 20 § 5, Pg. 20 Fig. 16) comprising the following steps: obtaining an infrared maritime ship data set; (Ye et al., Pg. 1 Abstract, Pg. 5 § 3 - § 3.1, Pg. 11 § 4 - Pg. 13 § 4.2.2, Pg. 13 Fig. 9) dividing the infrared maritime ship data set into an infrared ship detection training set and a verification set according to a ratio, (Ye et al., Pg. 11 § 4 - Pg. 13 § 4.2.2, Pg. 19 First-Full Paragraph) performing data enhancement preprocessing on the infrared maritime ship data set, (Ye et al., Pg. 5 § 3.1 ¶ 1, Pg. 11 § 4 - Pg. 13 § 4.2.2, Pg. 13 Fig. 9) wherein the data enhancement preprocessing includes Mosaic image enhancement; (Ye et al., Pg. 5 § 3.1 ¶ 1, Pg. 11 § 4 - Pg. 13 § 4.2.2, Pg. 13 Fig. 9) reforming a YOLO network structure, and obtaining an infrared ship target detection model by introducing an attention mechanism; (Ye et al., Pg. 1 Abstract, Pg. 2 First-Full Paragraph - Second-Full Paragraph, Pg. 5 § 3 - Pg. 6 § 3.2 ¶ 1, Pg. 7 § 3.3 - Pg. 9 § 3.4 ¶ 1, Pg. 9 Figs. 4 & 5, Pg. 20 § 5) training and verifying the infrared ship target detection model based on the infrared maritime ship data set to obtain an infrared ship target detection model; (Ye et al., Pg. 1 Abstract, Pg. 2 First-Full Paragraph - Second-Full Paragraph, Pg. 11 § 4 - Pg. 13 § 4.2.2, Pg. 15 Figs. 10 & 11, Pg. 16 Fig. 12, Pg. 20 § 5, Pg. 20 Fig. 16) and detecting a maritime ship based on the trained infrared ship target detection model. (Ye et al., Pg. 1 Abstract, Pg. 2 First-Full Paragraph - Second-Full Paragraph, Pg. 11 § 4 - Pg. 13 § 4.2.2, Pg. 15 Figs. 10 & 11, Pg. 16 Fig. 12, Pg. 20 § 5, Pg. 20 Fig. 16) Ye et al. fail to disclose explicitly converting the divided data set in Visual Object Classes (VOC) format into a data set in YOLO format, and wherein the data enhancement preprocessing includes left-right flipping and image zooming. Pertaining to analogous art, Chen et al. disclose performing data enhancement preprocessing on the infrared data set, (Chen et al., Abstract, Pg. 1 ¶ 0018 - Pg. 2 ¶ 0025, Pg. 2 ¶ 0035 and 0053, Pg. 3 ¶ 0066 - Pg. 4 ¶ 0073) wherein the data enhancement preprocessing includes left-right flipping, image zooming, and Mosaic image enhancement. (Chen et al., Pg. 2 ¶ 0022 - 0025 and 0035, Pg. 3 ¶ 0058 and 0066 - 0067, Pg. 4 ¶ 0072 - 0073) Chen et al. fail to disclose explicitly converting the divided data set in Visual Object Classes (VOC) format into a data set in YOLO format. Pertaining to analogous art, Wu et al. disclose converting the divided data set in Visual Object Classes (VOC) format into a data set in YOLO format. (Wu et al., Pg. 3 § 2 ¶ 1 - 2, Pg. 4 Figs. 1 & 2, Pgs. 5 - 6 § 3.3) Wang et al. and Ye et al. are combinable because they are both directed towards image processing systems and methods that utilize machine learning models to detect objects in images. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Wang et al. with the teachings of Ye et al. This modification would have been prompted in order to enhance the base device of Wang et al. with the well-known and applicable technique Ye et al. applied to a comparable device. Training a target detection model to detect maritime ships based on an infrared maritime ship data set, as taught by Ye et al., would enhance the base device of Wang et al. by allowing for it to be employed in a wider variety of situations and/or utilized in an increased number and variety of related and applicable applications and/or environments, such as for the detection of target objects in infrared images, thereby improving its overall appeal, usefulness and marketability to potential end-users. This combination could be completed according to well-known techniques in the art and would likely yield predictable results, in that the target detection model of the base device of Wang et al. would be trained on an infrared maritime ship data set and subsequently utilized to detect maritime ships in order to increase the number and variety of applications, environments and/or situations in which it may be employed so as to improve its overall appeal, usefulness and marketability to potential end-users. In addition, Wang et al. in view of Ye et al. and Chen et al. are combinable because they are all directed towards image processing systems and methods that utilize machine learning models to detect objects and/or targets in images. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the combined teachings of Wang et al. in view of Ye et al. with the teachings of Chen et al. This modification would have been prompted in order to enhance the combined base device of Wang et al. in view of Ye et al. with the well-known and applicable technique Chen et al. applied to a comparable device. Performing data enhancement preprocessing including left-right flipping, image zooming, and Mosaic image enhancement on the infrared data set, as taught by Chen et al., would enhance the combined base device by greatly increasing the number, quality and diversity of samples in the infrared data set in order to enhance the robustness and generalization ability of the infrared ship target detection model and reduce the chances and/or amount of overfitting occurring during training due to an insufficient number of samples in the infrared data set, as taught and suggested by Chen et al., see at least the abstract, page 2 paragraph 0040, page 3 paragraph 0071 - page 4 paragraph 0072 and page 4 paragraphs 0080 - 0081 of Chen et al. Furthermore, this modification would have been prompted by the teachings and suggestions of Ye et al. that the mosaic data augmentation method is used to increase the number of small objects in the infrared maritime ship data set which includes as few as 761 images of one category of detection target, see at least page 11 section 4.1 - page 13 line 9 and page 13 figure 9 of Ye et al. This combination could be completed according to well-known techniques in the art and would likely yield predictable results, in that data enhancement preprocessing including left-right flipping, image zooming, and Mosaic image enhancement would be performed on the infrared data set in order to enhance the robustness and generalization ability of the infrared ship target detection model of the combined base device and reduce the chances and/or amount of overfitting occurring during training by increasing the number, quality and diversity of samples in the infrared data set. Additionally, Wang et al. in view of Ye et al. in view of Chen et al. and Wu et al. are combinable because they are all directed towards image processing systems and methods that utilize machine learning models to detect objects and/or targets in images. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the combined teachings of Wang et al. in view of Ye et al. in view of Chen et al. with the teachings of Wu et al. This modification would have been prompted in order to enhance the combined base device of Wang et al. in view of Ye et al. in view of Chen et al. with the well-known and applicable technique Wu et al. applied to a comparable device. Converting the divided data set in Visual Object Classes (VOC) format into a data set in YOLO format, as taught by Wu et al., would enhance the combined base device by ensuring that all of the samples in the data set to be used for training, verifying and testing the infrared ship target detection model are in the correct format, by allowing for samples in VOC format to be collected and added to the data set for use by the infrared ship target detection model, and by facilitating evaluation of performance of the infrared ship target detection model against other models that utilize data sets in VOC formats so as to help increase the number and diversity of samples in the data set and simplify the process of collecting samples for use by the combined base device, since samples in VOC format would be able to be utilized, and to enable accurate performance evaluation of the trained infrared ship target detection model against benchmark data sets. Furthermore, this modification would have been prompted by the teachings and suggestions of Wang et al. that images in their dataset are labeled using Labellmg, see at least page 769 section subsection A paragraph 1 of Wang et al., and by the teachings of Wu et al. that their images were “labeled via labellmg1 software, which generates annotations in Pascal VOC XML [5] format, followed by their conversion into YOLO [6] format”, see at least page 3 section 2 paragraph 1 and pages 5 - 6 section 3.3 of Wu et al. Moreover, this modification would have been prompted by the teachings and suggestions of Wu et al. that Pascal VOC format and YOLO format are extensively employed in object detection tasks, see at least page 3 section 2 paragraphs 1 - 2 and pages 5 - 6 section 3.3 of Wu et al. This combination could be completed according to well-known techniques in the art and would likely yield predictable results, in that the divided data set in Visual Object Classes (VOC) format would be converted into a data set in YOLO format so as to ensure that all of the samples in the data set are in the correct format for use by the combined base device, simplify the process of collecting a sufficient number of samples for use by the combined base device, and enable accurate performance evaluation of the trained infrared ship target detection model against benchmark data sets in different formats, such as VOC format. Therefore, it would have been obvious to combine Wang et al. with Ye et al., Chen et al. and Wu et al. to obtain the invention as specified in claim 1.
- With regards to claim 3, Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. disclose the method for detecting an infrared ship target based on an improved YOLOv7 according to claim 1, wherein a process of reforming the YOLOv7 network structure based on the MobileNetv3 network and the bidirectional weighted feature pyramid network comprises: replacing a backbone feature extraction network in the YOLOv7 network structure with the MobileNetv3 network, (Wang et al., Pg. 766 Abstract, Pg. 766 § I ¶ 3, Pg. 767 § II - Pg. 768 Subsection B, Pg. 769 Subsection B - Pg. 770 Section “Discussion and Analysis”) and replacing a feature fusion network in the YOLOv7 network structure with the bidirectional weighted feature pyramid network. (Wang et al., Pg. 766 Abstract, Pg. 766 § I ¶ 3, Pg. 767 § II - Pg. 768 Subsection B, Pg. 769 Subsection B - Pg. 770 Section “Discussion and Analysis”)
- With regards to claim 4, Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. disclose the method for detecting an infrared ship target based on an improved YOLOv7 according to claim 3, wherein the MobileNetv3 network combines a depthwise separable convolution structure and an inverted residual structure, (Wang et al., Pg. 766 Abstract, Pgs. 767 - 768 Subsection A, Pg. 767 Figs. 1 & 2) and is integrated into a channel attention mechanism network; (Wang et al., Pg. 766 Abstract, Pgs. 767 - 768 Subsection A, Pg. 767 Figs. 1 & 2) wherein, the depthwise separable convolution structure comprises a depthwise convolution and a pointwise convolution. (Wang et al., Pg. 766 Abstract, Pgs. 767 - 768 Subsection A, Pg. 767 Figs. 1 & 2)
- With regards to claim 5, Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. disclose the method for detecting an infrared ship target based on an improved YOLOv7 according to claim 3, wherein the bidirectional weighted feature pyramid network increases a feature image weight, (Wang et al., Pg. 766 Abstract, Pg. 768 Subsection B, Pg. 768 Fig. 3) introduces a residual strategy, (Wang et al., Pg. 766 Abstract, Pg. 768 Subsection B, Pg. 768 Fig. 3) deletes nodes with low contribution, (Wang et al., Pg. 766 Abstract, Pg. 768 Subsection B, Pg. 768 Fig. 3) and adds an intermediate feature channel. (Wang et al., Pg. 766 Abstract, Pg. 768 Subsection B, Pg. 768 Fig. 3)
- With regards to claim 6, Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. disclose the method for detecting an infrared ship target based on an improved YOLOv7 according to claim 5, wherein a process of increasing the feature image weight comprises: the bidirectional weighted feature pyramid network automatically learns weight parameters of respective input feature layers, (Wang et al., Pg. 766 Abstract, Pg. 768 Subsection B, Pg. 768 Fig. 3) and then performs a weighted feature fusion on the input feature layers having different resolutions by using corresponding weight parameters and performs an output; (Wang et al., Pg. 766 Abstract, Pg. 768 Subsection B, Pg. 768 Fig. 3) wherein the bidirectional weighted feature pyramid network adds a jump connection between each input feature layer and an output feature layer in a same layer. (Wang et al., Pg. 766 Abstract, Pg. 768 Subsection B, Pg. 768 Fig. 3)
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Wang et al., “An improved YOLOv7 method for vehicle detection in traffic scenes”, IEEE, 35th Chinese Control and Decision Conference (CCDC), May 2023, pages 766 - 771 in view of Jing Ye, Zhaoyu Yuan, Cheng Qian, and Xiaoqiong Li, “CAA-YOLO: Combined-Attention-Augmented YOLO for Infrared Ocean Ships Detection”, Sensors, Vol. 22, Issue 10, May 2022, pages 1 - 23, herein referred to as “Ye et al.”, in view of Chen et al. U.S. Publication No. 2023/0188671 A1 in view of Zhen-wei Wu, Ming-hao Liu, Cheng-xiu Sun and Xin-fa Wang, “A dataset of tomato fruits images for object detection in the complex lighting environment of plant factories”, Data in Brief, Vol. 48, Article 109291, June 2023, pages 1 - 6, herein referred to as “Wu et al.”, as applied to claim 6 above, and further in view of Mingxing Tan, Ruoming Pang, and Quoc Le, “EfficientDet: Scalable and Efficient Object Detection”, IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020, pages 10778 - 10787, herein referred to as “Tan et al.”.
- With regards to claim 7, Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. disclose the method for detecting an infrared ship target based on an improved YOLOv7 according to claim 6. Wang et al. fail to disclose explicitly wherein calculation formulas of the weighted feature fusion is as follows:
PNG
media_image1.png
113
335
media_image1.png
Greyscale
wherein
p
i
t
d
and
p
i
o
u
t
represent intermediate transition features of an i-layer on a top-down path and final output features of an i-layer on a down-top path;
p
i
i
n
represents an input feature of the i-th layer;
p
i
+
1
i
n
represents an input feature of an (i+1)-th layer; and
p
i
-
1
o
u
t
represents a final output feature of an (i-1)-th layer; w1 and w2 respectively represent the weight parameters for calculating an input of a current layer and an input of a next layer of the intermediate transition features; w1′, w2′ and w3′ respectively represent a weight of the input of the current layer, a weight of an output of a transition unit of the current layer and a weight of an output of a previous layer, and ∈ value is 0.0001, and Conv stands for a convolution operation on a whole calculation result. Pertaining to analogous art, Tan et al. disclose wherein calculation formulas of the weighted feature fusion is as follows:
PNG
media_image1.png
113
335
media_image1.png
Greyscale
wherein
p
i
t
d
and
p
i
o
u
t
represent intermediate transition features of an i-layer on a top-down path and final output features of an i-layer on a down-top path;
p
i
i
n
represents an input feature of the i-th layer;
p
i
+
1
i
n
represents an input feature of an (i+1)-th layer; and
p
i
-
1
o
u
t
represents a final output feature of an (i-1)-th layer; w1 and w2 respectively represent the weight parameters for calculating an input of a current layer and an input of a next layer of the intermediate transition features; w1′, w2′ and w3′ respectively represent a weight of the input of the current layer, a weight of an output of a transition unit of the current layer and a weight of an output of a previous layer, and ∈ value is 0.0001, and Conv stands for a convolution operation on a whole calculation result. (Tan et al., Pgs. 10779 - 10781 § 3.3) Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. and Tan et al. are combinable because they are all directed towards image processing systems and methods that utilize machine learning models to detect objects and/or targets in images. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the combined teachings of Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. with the teachings of Tan et al. This modification would have been prompted in order to substitute the undisclosed weighted feature fusion formulas of Wang et al. for the weighted feature fusion formulas of Tan et al. The weighted feature fusion formulas of Tan et al. could be substituted in place of the undisclosed weighted feature fusion formulas of Wang et al. utilizing well-known techniques in the art and would likely yield predictable results, in that, in the combination, the weighted feature fusion formulas of Tan et al. would be utilized to fuse the feature information of the different feature layers of the combined base device. This combination could be completed according to well-known techniques in the art and would likely yield predictable results, in that the weighted feature fusion formulas of Tan et al. would be utilized to fuse the feature information of the different feature layers of the combined base device. Therefore, it would have been obvious to combine Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. with Tan et al. to obtain the invention as specified in claim 7.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Wang et al., “An improved YOLOv7 method for vehicle detection in traffic scenes”, IEEE, 35th Chinese Control and Decision Conference (CCDC), May 2023, pages 766 - 771 in view of Jing Ye, Zhaoyu Yuan, Cheng Qian, and Xiaoqiong Li, “CAA-YOLO: Combined-Attention-Augmented YOLO for Infrared Ocean Ships Detection”, Sensors, Vol. 22, Issue 10, May 2022, pages 1 - 23, herein referred to as “Ye et al.”, in view of Chen et al. U.S. Publication No. 2023/0188671 A1 in view of Zhen-wei Wu, Ming-hao Liu, Cheng-xiu Sun and Xin-fa Wang, “A dataset of tomato fruits images for object detection in the complex lighting environment of plant factories”, Data in Brief, Vol. 48, Article 109291, June 2023, pages 1 - 6, herein referred to as “Wu et al.”, as applied to claim 1 above, and further in view of Li Xiangrong and Sun Lihui, “Multiscale Infrared Target Detection Based on Attention Mechanism”, “Infrared Technology, Vol. 45, Issue 7, July 2023, pages 746 - 754, herein referred to as “Xiangrong et al.”. The Examiner notes that the citations to Xiangrong et al. correspond to the previously provided machine translation.
- With regards to claim 8, Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. disclose the method for detecting an infrared ship target based on an improved YOLOv7 according to claim 1, wherein the attention mechanism is an SENet (Squeeze-and-Excitation Networks) structure. (Wang et al., Pg. 767 § II - Pg. 768 Subsection B) Wang et al. fail to disclose expressly an SENet structure with a soft attention mechanism, and the SENet structure is used to extract an importance degree of each feature channel by an active learning method, then give the each feature channel different weights, and finally perform a filtration processing for features in a detection task based on a weight of the each feature channel. Pertaining to analogous art, Xiangrong et al. disclose wherein the attention mechanism is an SENet (Squeeze-and-Excitation Networks) structure with a soft attention mechanism, (Xiangrong et al., Pg. 747 Left-Hand Column Second-Full Paragraph - § 1.1, Pg. 747 Fig. 1, Pg. 749 § 1.5, Pg. 749 Fig. 4) and the SENet structure is used to extract an importance degree of each feature channel by an active learning method, (Xiangrong et al., Pg. 747 Left-Hand Column Second-Full Paragraph - § 1.1, Pg. 747 Fig. 1, Pg. 749 § 1.5, Pg. 749 Fig. 4) then give the each feature channel different weights, (Xiangrong et al., Pg. 747 Left-Hand Column Second-Full Paragraph - § 1.1, Pg. 747 Fig. 1, Pg. 749 § 1.5, Pg. 749 Fig. 4) and finally perform a filtration processing for features in a detection task based on a weight of the each feature channel. (Xiangrong et al., Pg. 747 Left-Hand Column Second-Full Paragraph - § 1.1, Pg. 747 Fig. 1, Pg. 749 § 1.5, Pg. 749 Fig. 4) Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. and Xiangrong et al. are combinable because they are all directed towards image processing systems and methods that utilize machine learning models to detect objects and/or targets in images. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the combined teachings of Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. with the teachings of Xiangrong et al. This modification would have been prompted in order to substitute the SE attention mechanism of Wang et al. for the Squeeze-and-Excitation Networks (SENet) channel attention mechanism of Xiangrong et al. The SENet channel attention mechanism of Xiangrong et al. could be substituted in place of the SE attention mechanism of Wang et al. utilizing well-known techniques in the art and would likely yield predictable results, in that, in the combination, the SENet channel attention mechanism of Xiangrong et al. would be utilized to enable the combined base device to focus on and strengthen the representational power of useful information and features. This combination could be completed according to well-known techniques in the art and would likely yield predictable results, in that the SENet channel attention mechanism of Xiangrong et al. would be utilized to enable the combined base device to focus on and strengthen the representational power of useful information and features. Therefore, it would have been obvious to combine Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. with Xiangrong et al. to obtain the invention as specified in claim 8.
Claim 10 is rejected under 35 U.S.C. 103 as being unpatentable over Wang et al., “An improved YOLOv7 method for vehicle detection in traffic scenes”, IEEE, 35th Chinese Control and Decision Conference (CCDC), May 2023, pages 766 - 771 in view of Jing Ye, Zhaoyu Yuan, Cheng Qian, and Xiaoqiong Li, “CAA-YOLO: Combined-Attention-Augmented YOLO for Infrared Ocean Ships Detection”, Sensors, Vol. 22, Issue 10, May 2022, pages 1 - 23, herein referred to as “Ye et al.”, in view of Chen et al. U.S. Publication No. 2023/0188671 A1 in view of Zhen-wei Wu, Ming-hao Liu, Cheng-xiu Sun and Xin-fa Wang, “A dataset of tomato fruits images for object detection in the complex lighting environment of plant factories”, Data in Brief, Vol. 48, Article 109291, June 2023, pages 1 - 6, herein referred to as “Wu et al.”, as applied to claim 1 above, and further in view of Zheng et al., “A lightweight ship target detection model based on improved YOLOv5 algorithm”, PLoS ONE, 18(4), Apr. 2023, pages 1 - 23.
- With regards to claim 10, Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. disclose the method for detecting an infrared ship target based on an improved YOLOv7 according to claim 1, wherein a process of training and verifying the infrared ship target detection model comprises: setting an initial learning rate and initial iterations of the target detection model, (Wang et al., Pg. 769 § III - Subsection B) and adaptively adjusting a scaling of the training set, the verification set and the test set based on a preset input image size; (Wang et al., Pg. 767 Fig. 1, Pg. 768 Subsection B, Pg. 769 § III - Subsection B [“the resolution of the input image is set to 618x618px,”]) verifying the target detection model based on the training set and the verification set after adaptively adjusting the scaling of the training set and the verification set based on the preset input image size, (Wang et al., Pg. 767 Fig. 1, Pg. 769 § III - Pg. 770 Section “Discussion and Analysis”, Pg. 770 Fig. 7) and adjusting the initial learning rate and the initial iterations, so as to obtain a target learning rate and target iterations (Wang et al., Pg. 769 § III - Pg. 770 Section “Discussion and Analysis”) and further obtain the target detection model trained; (Wang et al., Pg. 769 § III - Pg. 770 Section “Discussion and Analysis”, Pg. 770 Fig. 7) and finally, testing the target detection model trained based on the test set after adaptively adjusting the scaling of the test set based on the preset input image size. (Wang et al., Pg. 767 Fig. 1, Pg. 768 Subsection B, Pg. 769 § III - Pg. 770 Section “Discussion and Analysis”, Pg. 770 Fig. 7) Wang et al. fail to disclose explicitly the infrared ship target detection model; cross-verifying an average accuracy change and loss change trend of the infrared ship target detection model based on the training set and the verification set, and adjusting the initial learning rate and the initial iterations until the average accuracy change and the loss change tend to be stable. Pertaining to analogous art, Ye et al. disclose wherein a process of training and verifying the infrared ship target detection model comprises: setting an initial learning rate and initial iterations of the infrared ship target detection model, (Ye et al., Pg. 11 § 4 - Pg. 13 § 4.2.2, Pg. 19 First-Full Paragraph - Third-Full Paragraph) and adaptively adjusting a scaling of the training set, the verification set and the test set based on a preset input image size; (Ye et al., Pg. 11 § 4 - Pg. 13 § 4.2.2) verifying the infrared ship target detection model based on the training set and the verification set after adaptively adjusting the scaling of the training set and the verification set based on the preset input image size, (Ye et al., Pg. 1 Abstract, Pg. 2 First-Full Paragraph - Second-Full Paragraph, Pg. 11 § 4 - Pg. 13 § 4.2.2, Pg. 15 Figs. 10 & 11, Pg. 16 Fig. 12, Pg. 17 First-Full Paragraph - Pg. 19 Third-Full Paragraph, Pg. 20 § 5, Pg. 20 Fig. 16) and adjusting the initial learning rate and the initial iterations, so as to obtain a target learning rate and target iterations, (Ye et al., Pg. 11 § 4 - Pg. 13 § 4.2.2, Pg. 19 First-Full Paragraph - Third-Full Paragraph) and further obtain the infrared ship target detection model trained; (Ye et al., Pg. 1 Abstract, Pg. 2 First-Full Paragraph - Second-Full Paragraph, Pg. 11 § 4 - Pg. 13 § 4.2.2, Pg. 15 Figs. 10 & 11, Pg. 16 Fig. 12, Pg. 19 First-Full Paragraph - Third-Full Paragraph, Pg. 20 § 5, Pg. 20 Fig. 16) and finally, testing the infrared ship target detection model trained based on the test set after adaptively adjusting the scaling of the test set based on the preset input image size. (Ye et al., Pg. 1 Abstract, Pg. 2 First-Full Paragraph - Second-Full Paragraph, Pg. 11 § 4 - Pg. 13 § 4.2.2, Pg. 15 Figs. 10 & 11, Pg. 16 Fig. 12, Pg. 19 First-Full Paragraph - Third-Full Paragraph, Pg. 20 § 5, Pg. 20 Fig. 16) Ye et al. fail to disclose explicitly cross-verifying an average accuracy change and loss change trend of the infrared ship target detection model based on the training set and the verification set, and adjusting the initial learning rate and the initial iterations until the average accuracy change and the loss change tend to be stable. Pertaining to analogous art, Zheng et al. disclose wherein a process of training and verifying the infrared ship target detection model comprises: setting an initial learning rate and initial iterations of the infrared ship target detection model; (Zheng et al., Pg. 1 Abstract, Pg. 12 § 4.2 - Pg. 15 First-Full Paragraph, Pg. 15 Fig. 11) cross-verifying an average accuracy change and loss change trend of the infrared ship target detection model based on the training set and the verification set, (Zheng et al., Pg. 13 § 4.3 - Pg. 16 Line 5, Pg. 15 Fig. 11) and adjusting the initial learning rate and the initial iterations until the average accuracy change and the loss change tend to be stable, so as to obtain a target learning rate and target iterations, (Zheng et al., Pg. 12 § 4.2 - Pg. 16 Line 5, Pg. 15 Fig. 11) and further obtain the infrared ship target detection model trained; (Zheng et al., Pg. 1 Abstract, Pg. 13 § 4.3 - Pg. 15 First-Full Paragraph, Pg. 15 Fig. 1, Pgs. 20 - 21 § 6) and finally, testing the infrared ship target detection model trained based on the test set. (Zheng et al., Pg. 1 Abstract, Pg. 13 § 4.3 - Pg. 19 § 5.2, Pg. 18 Fig. 16, Pgs. 20 - 21 § 6) Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. and Zheng et al. are combinable because they are all directed towards image processing systems and methods that utilize machine learning models to detect objects and/or targets in images. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the combined teachings of Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. with the teachings of Zheng et al. This modification would have been prompted in order to enhance the combined base device of Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. with the well-known and applicable technique Zheng et al. applied to a comparable device. Cross-verifying an average accuracy change and loss change trend based on the training set and the verification set and adjusting the initial learning rate and the initial iterations until the average accuracy change and the loss change tend to be stable, as taught by Zheng et al., would enhance the combined base device by improving its ability to effectively and reliably obtain a sufficiently trained target detection model exhibiting a high level of performance since the target detection model would be trained until a point where subsequent training does not yield any appreciable amount of improvement in performance. Furthermore, this modification would have been prompted by the teachings and suggestions of Wang et al. that their dataset is divided into training, validation and test sets and that a stochastic gradient descent optimization strategy is used for training, see at least page 769 section III - subsection B of Wang et al. Moreover, this modification would have been prompted by the teachings and suggestions of Ye et al. that their data set is split into training and validation sets and that a stochastic gradient descent optimization and cosine learning rate decay strategy is used for training, see at least pages 12 - 13 section 4.2.1 of Ye et al. This combination could be completed according to well-known techniques in the art and would likely yield predictable results, in that an average accuracy change and loss change trend would be cross-verified based on the training set and the verification set and that the initial learning rate and the initial iterations would be adjusted until the average accuracy change and the loss change tend to be stable so as to enhance the ability of the combined base device to effectively and reliably obtain a sufficiently trained target detection model exhibiting a high level of performance. Therefore, it would have been obvious to combine Wang et al. in view of Ye et al. in view of Chen et al. in view of Wu et al. with Zheng et al. to obtain the invention as specified in claim 10.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ERIC RUSH whose telephone number is (571) 270-3017. The examiner can normally be reached 9am - 5pm Monday - Friday.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Andrew Bee can be reached at (571) 270 - 5183. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/ERIC RUSH/Primary Examiner, Art Unit 2677
Prosecution Insights