DETAILED ACTIONS
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 .
Priority
Acknowledgment is made of applicant’s claim this application being in benefit of Provisional Patent Application No. 63/580,659 filed on September 5th, 2023.
Information Disclosure Statement
The information disclosure statement (“IDS”) filed on 12/23/2024 was reviewed and the listed references were noted.
Drawings
The 18-page drawings have been considered and placed on record in the file.
Status of Claims
Claims 1-20 are pending.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1-3, 8-11, and 16-19 are rejected under 35 U.S.C. 103 as being unpatentable over Caesar (US 2020/0272854 A1), hereinafter referred to as Caesar, in view of Zhou et al., (US 11,810,371 B1), hereinafter referred to as Zhou.
Claim 1
Caesar discloses a method for selecting one or more Regions of Interest (RoIs) for human annotations (Caesar, Fig. 18) comprising:
obtaining sensor data generated by one or more sensors (Caesar, [0014], “the plurality of different machine learning models includes a plurality of different neural networks tuned by training data samples provided by different types of sensors”) of a vehicle (Caesar, [0070], “this document describes technologies applicable to any vehicles that have one or more autonomous capabilities including fully autonomous vehicles”);
applying at least one class-agnostic heuristic function to the sensor data to determine a presence and an approximate position of one or more objects in an RoI of the sensor data (Caesar, [0100], “the perception module 402 includes an object detector that detects and labels objects. An example object detector is a convolutional neural network (CNN). In an embodiment, the object detector can output an image or point cloud that includes bounding boxes surrounding the detected objects, labels for the objects and bounding box scores that indicate a degree of confidence that the objects detected were correctly labeled. In an embodiment, the bounding box score can be in the range [0.0 1.0], where 0.0 indicates an incorrectly labeled object, 1.0 indicates a correctly labeled object and values in between indicate a degree of confidence that the object was correctly labeled.”, the detection determines the presence or existence of the object in the data such as image and the bounding box determines the position of the one or more object in the image, the bounding box is analogous to RoI, as it is being further processed to be further labeled by an annotator);
selecting one or more RoIs having proposed annotations for the one or more objects (Caesar[0148], “Process 1800 continues by generating an ensemble score variance from sets of predictions output from multiple machine learning models (1804). For example, N machine learning (ML) models (e.g., 3 CNNs) are trained with training data (e.g., training images). The “ensemble” of machine learning models generates N sets of predictions which are fed into a pairwise comparator module. The pairwise comparator module computes the pairwise agreement between the N sets of predictions to generate N ensemble scores (e.g., set of N mAP values), where each set of predictions, in turn, is the “ground truth” for the other sets of predictions on a box level. An ensemble score variance generator then computes an ensemble score variance which is a measure of the difference between the N ensemble scores. When selecting bounding boxes for annotation, the active learning system will select data samples with the highest variance. In an alternative embodiment, cross-modal ensembles are used where a measure of agreement is made between object detections output by multiple different sensors (e.g., RADAR, LiDAR and camera).”) for refinement by an annotator (Caesar, [0132], “the active learning system 1400 automatically queries the annotator 1406 for labels. Data samples 1401 (e.g., images, point clouds) include bounding boxes with known labels and bounding boxes with unknown labels. The labeled bounding boxes can be correctly labeled or incorrectly labeled. The active learning system 1400 automatically selects a subset of bounding boxes with unknown and/or incorrect labels for manual labeling (annotating) by the annotator 1406.”); and
outputting the one or more selected RoIs (Caesar, Fig. 14, step 1405, the selected data samples or bounding box is output or sent to the human annotator, [0132], “The active learning system 1400 automatically selects a subset of bounding boxes with unknown and/or incorrect labels for manual labeling (annotating) by the annotator 1406.”).
Caesar does not explicitly disclose applying at least one class-agnostic heuristic function to the sensor data to determine a presence and an approximate position of one or more objects in an RoI of the sensor data.
However, Zhou teaches applying at least one class-agnostic heuristic function to the sensor data to determine a presence and an approximate position of one or more objects in an RoI of the sensor data (Zhou, Col. 6, lines 14-19, “, the encoded information may be transformed to a list of class-agnostic candidate objects (e.g., region proposals) using a region proposal network (RPN) (e.g., a neural network previously trained to generate region proposals from an image). The objects from the region proposals are pooled to region-of interest (ROI) features”, “ These ROI features may be fed to a second neural network that includes a classifier and a regressor head. A refined list of objects defined by class labels and bounding box coordinates is obtained from the classifier and the regressor head, respectively.”).
Caesar and Zhou are both considered to be analogous to the claimed invention because they are in the same field of object detection for autonomous vehicle. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified the image signal processor as taught by Caesar to incorporate the teachings of Zhou of applying at least one class-agnostic heuristic function to the sensor data to determine a presence and an approximate position of one or more objects in an RoI of the sensor data. Such a modification is the result of combining prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been to utilize fewer computations than conventional technique (Zhou, Col. 12, lines 5-6).
Claim 2
The combination of Caesar in view of Zhou discloses the method of claim 1 (Caesar, Fig. 18), further comprising:
applying a machine learning model to the sensor data to generate predicted annotations and one or more proposed RoIs (Caesar, [0100], “the perception module 402 includes an object detector that detects and labels objects. An example object detector is a convolutional neural network (CNN). In an embodiment, the object detector can output an image or point cloud that includes bounding boxes surrounding the detected objects, labels for the objects and bounding box scores that indicate a degree of confidence that the objects detected were correctly labeled. In an embodiment, the bounding box score can be in the range [0.0 1.0], where 0.0 indicates an incorrectly labeled object, 1.0 indicates a correctly labeled object and values in between indicate a degree of confidence that the object was correctly labeled.”, the detection determines the presence or existence of the object in the data such as image and the bounding box determines the position of the one or more object in the image, the bounding box is analogous to RoI, as it is being further processed to be further labeled by an annotator); and
analyzing the predicted annotations to generate the proposed annotations and to selectively refine, prior to outputting, the one or more proposed RoIs (Caesar[0148], “Process 1800 continues by generating an ensemble score variance from sets of predictions output from multiple machine learning models (1804). For example, N machine learning (ML) models (e.g., 3 CNNs) are trained with training data (e.g., training images). The “ensemble” of machine learning models generates N sets of predictions which are fed into a pairwise comparator module. The pairwise comparator module computes the pairwise agreement between the N sets of predictions to generate N ensemble scores (e.g., set of N mAP values), where each set of predictions, in turn, is the “ground truth” for the other sets of predictions on a box level. An ensemble score variance generator then computes an ensemble score variance which is a measure of the difference between the N ensemble scores. When selecting bounding boxes for annotation, the active learning system will select data samples with the highest variance. In an alternative embodiment, cross-modal ensembles are used where a measure of agreement is made between object detections output by multiple different sensors (e.g., RADAR, LiDAR and camera).”).
Claim 3
The combination of Caesar in view of Zhou discloses the method of claim 1 (Caesar, Fig. 18), wherein applying the at least one class-agnostic heuristic function (Zhou, Col. 6, lines 14-19, “, the encoded information may be transformed to a list of class-agnostic candidate objects (e.g., region proposals) using a region proposal network (RPN) (e.g., a neural network previously trained to generate region proposals from an image). The objects from the region proposals are pooled to region-of interest (ROI) features”, “ These ROI features may be fed to a second neural network that includes a classifier and a regressor head. A refined list of objects defined by class labels and bounding box coordinates is obtained from the classifier and the regressor head, respectively.”) comprises:
determining the presence and the approximate position of the one or more objects (Zhou, Col. 6, lines 14-19, “, the encoded information may be transformed to a list of class-agnostic candidate objects (e.g., region proposals) using a region proposal network (RPN) (e.g., a neural network previously trained to generate region proposals from an image). The objects from the region proposals are pooled to region-of interest (ROI) features”, “ These ROI features may be fed to a second neural network that includes a classifier and a regressor head. A refined list of objects defined by class labels and bounding box coordinates is obtained from the classifier and the regressor head, respectively.”) using a corresponding High Definition (HD) map (Zhou, Abstract, “Synthetic data may be generated by populating a high-definition (HD) map with an object. The populated HD map can be projected to a two-dimensional camera view image depicting the object. The HD map and object detection data generated from image (e.g., identifying lane/object locations within the image) can be used to train the model to identify HD map locations of an autonomous vehicle capturing images as it travels, as well as various objects detected from those captured images. Subsequently, a new image may be processed using object detection techniques to detect lane/object locations within the image. An HD map and the detected lane/object data may be provided to the model, which in turn identifies, on the HD map, the locations of the vehicle and the various objects.”).
The proposed combination as well as the motivation for combining the Caesar and Zhou references presented in the rejection of Claim 1, apply to Claim 3 and are incorporated herein by reference. Thus, the method recited in Claim 3 is met by Caesar and Zhou.
Claim 8
The combination of Caesar in view of Zhou discloses the method of claim 1 (Caesar, Fig. 18), wherein outputting the one or more selected RoIs (Caesar, Fig. 14, step 1405) comprises:
sending the one or more selected RoIs via an interface used by the annotator (Caesar, Fig. 14, step 1405, the selected data samples or bounding box is output or sent to the human annotator, [0132], “The active learning system 1400 automatically selects a subset of bounding boxes with unknown and/or incorrect labels for manual labeling (annotating) by the annotator 1406.”, [0078], “The communication interfaces 140 transmit data collected from sensors 121 or other data related to the operation of AV 100 to the remotely located database 134. In an embodiment, communication interfaces 140 transmit information that relates to teleoperations to the AV 100. In some embodiments, the AV 100 communicates with other remote (e.g., “cloud”) servers 136.”, [0132], “the active learning system 1400 automatically queries the annotator 1406 for labels”).
Claims 9-11 and 16 are rejected for similar reasons as those described in claims 1-3, and 8. The additional elements in Claims 9-11 and 16 (the combination of Caesar in view of Zhou) discloses includes: an apparatus (Caesar, Fig. 3) for selecting one or more Regions of Interest (RoIs) for human annotations (Caesar, Fig. 18), the apparatus comprising: a memory for storing sensor data (Caesar, Fig. 3, main memory 306) and processing circuitry in communication with the memory (Caesar, Fig. 3, processor 304 and main memory 306). The proposed combination as well as the motivation for combining the Caesar and Zhou references presented in the rejection of Claim 1, apply to Claims 9-11 and 16 and are incorporated herein by reference. Thus, the apparatus recited in Claims 9-11 and 16 is met by Caesar and Zhou.
Claims 17-19 are rejected for similar reasons as those described in claims 1-3. The additional elements in Claims 17-19 (the combination of Caesar in view of Zhou) discloses includes: non-transitory computer-readable storage media having instructions encoded thereon, the instructions configured to cause processing circuitry (Caesar, Fig. 3, processor 304 and main memory 306). The proposed combination as well as the motivation for combining the Caesar and Zhou references presented in the rejection of Claim 1, apply to Claims 17-19 and are incorporated herein by reference. Thus, the media recited in Claims 17-19 is met by Caesar and Zhou.
Claims 4, 12, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Caesar in view of Zhou, in further view of Bangalath et al., (US 2024/0203085 A1), hereinafter referred to as Bangalath.
Claim 4
The combination of Caesar in view of Zhou discloses the method of claim 1 (Caesar, Fig. 18), wherein applying the at least one class-agnostic heuristic function (Zhou, Col. 6, lines 14-19, “, the encoded information may be transformed to a list of class-agnostic candidate objects (e.g., region proposals) using a region proposal network (RPN) (e.g., a neural network previously trained to generate region proposals from an image). The objects from the region proposals are pooled to region-of interest (ROI) features”, “ These ROI features may be fed to a second neural network that includes a classifier and a regressor head. A refined list of objects defined by class labels and bounding box coordinates is obtained from the classifier and the regressor head, respectively.”).
The combination of Caesar in view of Zhou does not explicitly disclose calculating a respective objectness measure count for each of a plurality of frames of the sensor data, wherein the respective objectness measure count is indicative of the presence of objects within a corresponding frame and rejecting one or more of the plurality of frames based on the respective objectness measure count.
However, Bangalath teaches calculating a respective objectness measure count (Bangalath, [0097], “Class-agnostic object proposals 414 are generated from the MVIT using ‘all objects’ text query. The generated proposals are ranked using predicted objectness scores and the top 5 proposals (including bounding boxes 1202, 1204, 1206, 1208) per image are selected for RKD as shown in FIG. 12. Next, the CLIP image-encoder 416 and the present OVD detector is used to generate embeddings corresponding to these proposals which are then used for calculating the RKD loss in Eq. 3. To save computation load and increase the training efficiency, in one embodiment the class-agnostic proposals 414 and the corresponding CLIP region embeddings are computed offline and are loaded during training. ”) for each of a plurality of frames of the sensor data (Bangalath, [0047], “Object detection has been divided into two sub-tasks: localization, which involves determining the location of an object in an image (or video frame),”), wherein the respective objectness measure count is indicative of the presence of objects within a corresponding frame (Bangalath, Fig. 12, class-agnostic proposals, [0097], “Class-agnostic object proposals 414 are generated from the MVIT using ‘all objects’ text query. The generated proposals are ranked using predicted objectness scores and the top 5 proposals (including bounding boxes 1202, 1204, 1206, 1208) per image are selected for RKD as shown in FIG. 12.) and rejecting one or more of the plurality of frames based on the respective objectness measure count (Bangalath, [0097], “. The generated proposals are ranked using predicted objectness scores and the top 5 proposals (including bounding boxes 1202, 1204, 1206, 1208) per image are selected for RKD as shown in FIG. 12.”, the ones that are not in top 5 proposals based on scores are implied to be rejected).
Caesar, Zhou, and Bangalath are both considered to be analogous to the claimed invention because they are in the same field of object detection for autonomous vehicle. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified the image signal processor as taught by Caesar to incorporate the teachings of Bangalath of calculating a respective objectness measure count for each of a plurality of frames of the sensor data, wherein the respective objectness measure count is indicative of the presence of objects within a corresponding frame and rejecting one or more of the plurality of frames based on the respective objectness measure count. Such a modification is the result of combining prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been for more accurate detections.
Claim 12 is rejected for similar reasons as those described in claim 4. The additional elements in Claim 12 (the combination of Caesar in view of Zhou in further view of Bangalath) discloses includes: an apparatus (Caesar, Fig. 3) for selecting one or more Regions of Interest (RoIs) for human annotations (Caesar, Fig. 18), the apparatus comprising: a memory for storing sensor data (Caesar, Fig. 3, main memory 306) and processing circuitry in communication with the memory (Caesar, Fig. 3, processor 304 and main memory 306). The proposed combination as well as the motivation for combining the Caesar, Zhou, and Bangalath references presented in the rejection of Claim 4, apply to Claim 12 and are incorporated herein by reference. Thus, the apparatus recited in Claim 12 is met by Caesar, Zhou, and Bangalath.
Claim 20 is rejected for similar reasons as those described in claim 4. The additional elements in Claim 20 (the combination of Caesar in view of Zhou in further view of Bangalath) discloses includes: non-transitory computer-readable storage media having instructions encoded thereon, the instructions configured to cause processing circuitry (Caesar, Fig. 3, processor 304 and main memory 306). The proposed combination as well as the motivation for combining the Caesar, Zhou, and Bangalath references presented in the rejection of Claim 4, apply to Claim 20 and are incorporated herein by reference. Thus, the media recited in Claim 20 is met by Caesar, Zhou, and Bangalath.
Claims 5-6 and 13-14 are rejected under 35 U.S.C. 103 as being unpatentable over Caesar in view of Zhou, in further view of Filatov et al., “Any Motion Detector: Learning Class-agnostic Scene Dynamics from a Sequence of LiDAR Point Clouds” (April 2020), hereinafter referred to as Filatov.
Claim 5
The combination of Caesar in view of Zhou discloses the method of claim 1 (Caesar, Fig. 18), wherein applying the at least one class-agnostic heuristic function (Zhou, Col. 6, lines 14-19, “, the encoded information may be transformed to a list of class-agnostic candidate objects (e.g., region proposals) using a region proposal network (RPN) (e.g., a neural network previously trained to generate region proposals from an image). The objects from the region proposals are pooled to region-of interest (ROI) features”, “ These ROI features may be fed to a second neural network that includes a classifier and a regressor head. A refined list of objects defined by class labels and bounding box coordinates is obtained from the classifier and the regressor head, respectively.”).
The combination of Caesar in view of Zhou does not explicitly disclose detecting one or more areas with one or more moving objects to determine the approximate position of the one or more objects.
However, Filatov teaches detecting one or more areas with one or more moving objects (Filatov, Section IV.B, “To estimate the motion of dynamic objects in the scene we need to aggregate temporal context. Voxel Feature Encoder outputs tensor representations of consequent point clouds, but each of them is in its local coordinate system. If one naively aggregates such representations using RNN or 3D convolutions, the model would be forced to learn ego-motion of the observer implicitly.”) to determine the approximate position of the one or more objects (Filatov, Section V.A, “We argue that we can use 3D bounding boxes with assigned velocities to obtain segmentation and velocity grid by procedure described in Figure 4.”, Fig. 4, “We project 3D bounding boxes (a) and velocities onto the grid (b). Hue component indicates the velocity direction according to HSV color space. Saturation indicates the absolute velocity value. Refinement step (c) leaves only those cells that contain LiDAR points”).
Caesar, Zhou, and Filatov are both considered to be analogous to the claimed invention because they are in the same field of object detection for autonomous vehicle. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified the image signal processor as taught by Caesar to incorporate the teachings of Filatov of detecting one or more areas with one or more moving objects to determine the approximate position of the one or more objects.. Such a modification is the result of combining prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been to achieve high-performance and real-time inference with the end-to-end approach (Filatov, Section VI).
Claim 6
The combination of Caesar in view of Zhou in view of Filatov discloses the method of claim 5 (Caesar, Fig. 18), wherein detection the one or more areas (Filatov, Section IV.B, “To estimate the motion of dynamic objects in the scene we need to aggregate temporal context. Voxel Feature Encoder outputs tensor representations of consequent point clouds, but each of them is in its local coordinate system. If one naively aggregates such representations using RNN or 3D convolutions, the model would be forced to learn ego-motion of the observer implicitly.”) comprises:
analyzing changes in pixel intensity between two or more video frames to identify one or more motion edges (Filatov, Section I, “. To estimate object motion parameters one needs to incorporate temporal context into the object detection model and analyze several consecutive frames.”, Fig. 3, “RNN cell with ego-motion compensation layer. We take a hidden state and apply an affine transform to it using the odometry between the previous and current frames.”.
The proposed combination as well as the motivation for combining the Caesar, Zhou, and Filatov references presented in the rejection of Claim 5, apply to Claim 6 and are incorporated herein by reference. Thus, the method recited in Claim 6 is met by Caesar, Zhou, and Filatov.
Claims 13-14 are rejected for similar reasons as those described in claims 5-6. The additional elements in Claims 13-14 (the combination of Caesar in view of Zhou in further view of Filatov) discloses includes: an apparatus (Caesar, Fig. 3) for selecting one or more Regions of Interest (RoIs) for human annotations (Caesar, Fig. 18), the apparatus comprising: a memory for storing sensor data (Caesar, Fig. 3, main memory 306) and processing circuitry in communication with the memory (Caesar, Fig. 3, processor 304 and main memory 306). The proposed combination as well as the motivation for combining the Caesar, Zhou, and Filatov references presented in the rejection of Claim 5, apply to Claims 13-14 and are incorporated herein by reference. Thus, the apparatus recited in Claims 13-14 is met by Caesar, Zhou, and Filatov.
Claims 7 and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Caesar in view of Zhou, in further view of Rejeb Sfar et al., (US 20220189070 A1), hereinafter referred to as Rejeb Sfar.
Claim 7
The combination of Caesar in view of Zhou discloses the method of claim 1 (Caesar, Fig. 18), wherein applying the at least one class-agnostic heuristic function (Zhou, Col. 6, lines 14-19, “, the encoded information may be transformed to a list of class-agnostic candidate objects (e.g., region proposals) using a region proposal network (RPN) (e.g., a neural network previously trained to generate region proposals from an image). The objects from the region proposals are pooled to region-of interest (ROI) features”, “ These ROI features may be fed to a second neural network that includes a classifier and a regressor head. A refined list of objects defined by class labels and bounding box coordinates is obtained from the classifier and the regressor head, respectively.”).
The combination of Caesar in view of Zhou does not explicitly disclose determining shape of the one or more objects to determine the approximate position of the one or more objects.
However, Rejeb Sfar teaches determining shape of the one or more objects to determine the approximate position of the one or more objects (Rejeb Sfar, [0153], “The implementation allows example-based 3D object detection, which consists in localizing similar shapes/objects to a given point cloud example, in a larger point cloud. Localization is based on 3D bounding-boxes around detected objects. The implementation employs a deep learning approach for the detection. The assessment criteria for this implementation are as follows: [0154] Given as input a point cloud query (example) and a larger point cloud (search area), [0155] The implementation seeks to have as output 3D bounding-boxes around similar objects to the query, in the search area. However, it is not known in advance the number of objects to be detected. Note that the search area could not contain similar object to the selected query. [0156] The Example-based object detection should be class-agnostic, i.e., able to detect similar object to any object query.“. localizing the object determines the approximate position of the object).
Caesar, Zhou, and Rejeb Sfar are all considered to be analogous to the claimed invention because they are in the same field of object detection for autonomous vehicle. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified the image signal processor as taught by Caesar to incorporate the teachings of Rejeb Sfar of determining shape of the one or more objects to determine the approximate position of the one or more objects. Such a modification is the result of combining prior art elements according to known methods to yield predictable results. The motivation for the proposed modification would have been so it could be robust to noise, occlusion, point distribution and also invariant to different types of transformation (Rejeb Sfar, [0157]).
Claim 15 is rejected for similar reasons as those described in claim 7. The additional elements in Claim 15 (the combination of Caesar in view of Zhou in further view of Rejeb Sfar) discloses includes: an apparatus (Caesar, Fig. 3) for selecting one or more Regions of Interest (RoIs) for human annotations (Caesar, Fig. 18), the apparatus comprising: a memory for storing sensor data (Caesar, Fig. 3, main memory 306) and processing circuitry in communication with the memory (Caesar, Fig. 3, processor 304 and main memory 306). The proposed combination as well as the motivation for combining the Caesar, Zhou, and Rejeb Sfar references presented in the rejection of Claim 7, apply to Claim 15 and are incorporated herein by reference. Thus, the apparatus recited in Claim 15 met by Caesar, Zhou, and Rejeb Sfar.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to DENISE G ALFONSO whose telephone number is (571)272-1360. The examiner can normally be reached Monday - Friday 7:30 - 5:30.
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, Amandeep Saini can be reached at (571)272-3382. 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.
/DENISE G ALFONSO/Examiner, Art Unit 2662 /AMANDEEP SAINI/Supervisory Patent Examiner, Art Unit 2662
Prosecution Insights