DETAILED ACTION
Response to Amendment
The amendment filed on 27 June 2025 has been entered.
Claims 1-20 are pending.
Claims 1-7, 10-11, 13-14, 17, 20 are cancelled.
Claims 8-9, 12, 15-16, 18 are amended.
Claims 21-24 are new.
Claims 8-9, 12, 15-16, 18-19, 21-24 will be pending.
Applicant’s amendments to the Claims have overcome each and every rejection under 35 USC 101 previously set forth in the Non-Final Office Action mailed 27 December 2024.
Response to Arguments
Applicant’s remarks, regarding the rejections of claims under 35 USC 103, have been fully considered.
Applicant notes Claims 1-7, 10-11, 13-14, 17, and 20 are cancelled herein without prejudice or disclaimer, rendering the rejections thereof moot. With regard to the remaining claims, and without acceding to the Office Action's assertions and in the interest of compact prosecution, Applicant has herein amended independent Claim 8 (with similar amendments made to independent Claim 15) to recite, among other things:
A system, comprising:
a representation learning engine, configured to:
receive input data via at least one of a sensor data stream, software-generated telemetry, an internet-based data feed, a source of structured digital content, a source of unstructured digital data, and perform representation learning by:
performing self-supervised clustering of features extracted from the input data into a hierarchical set of symbols, using a neuro-symbolic generator, to generate a learned feature language; and encoding the input data into at least one latent vector representing at least one pattern in the input data, the at least one latent vector included in the learned feature language; and a decision learning engine, configured to:
generate, via a plurality of codelets, a plurality of sub-decisions including detection of at least one of spatial anomalies, temporal anomalies, or composite anomalies, selectively execute codelets from the plurality of codelets in response to traversing a graph structure and satisfying associated conditions, and store a representation of at least one of the at least one latent vector, the spatial anomalies, the temporal anomalies, the composite anomalies, or the at least one pattern in at least one of a semantic memory, an episodic memory, or a long-term memory, the system configured to (1) update the learned feature language based on newly-received input data and data retrieved from the at least one of a semantic memory, an episodic memory, or a long-term memory, and (2) detect in real-time, and respond to, new anomalies.
Applicant submits none of Eaton, Vu, and Seow, whether taken individually or in proper combination (not herein conceded) discloses or suggests all recitations of independent Claims 8 and 15 as herein amended.
Applicant submits none of Eaton, Vu, and Seow discloses or suggests "a decision learning engine, configured to: generate, via a plurality of codelets, a plurality of sub-decisions including detection of at least one of spatial anomalies, temporal anomalies, or composite anomalies,"' as recited by amended independent Claim 8 (with similar amendments made to independent Claim 15). Applicant submits Knight and Seow2 fail to remedy the aforementioned deficiencies of Eaton, Vu, and Seow.
Applicant’s arguments 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.
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 .
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 27 June 2025 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered and attached by the examiner.
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 8, 15, 21-24 are rejected under 35 U.S.C. 103 as being unpatentable over Cobb et al. (U.S. Pre-Grant Publication No. 20110052000, hereinafter ‘Cobb), in view of Nakahata et al. (NPL: "Anomaly detection with a moving camera using spatio-temporal codebooks", hereinafter 'Nakahata').
Regarding claim 8 and analogous claim 15, Cobb teaches A system, comprising ([0028] One embodiment of the invention is implemented as a program product for use with a computer system. The program(s) of the program product defines functions of the embodiments (including the methods described herein) and can be contained on a variety of computer-readable storage media. Examples of computer-readable storage media include (i) non-writable storage media (e.g., read-only memory devices within a computer such as CD-ROM or DVD-ROM disks readable by an optical media drive) on which information is permanently stored; (ii) writable storage media (e.g., floppy disks within a diskette drive or hard-disk drive) on which alterable information is stored.):
a representation learning engine, configured to ([0030] FIG. 1 illustrates components of a video analysis and behavior-recognition system 100, according to one embodiment of the invention.):
receive input data via at least one of a sensor data stream, software-generated telemetry, an internet-based data feed, a source of structured digital content, or a source of unstructured digital data ([0030] FIG. 1 illustrates components of a video analysis and behavior-recognition system 100, according to one embodiment of the invention. As shown, the behavior-recognition system 100 includes a video receive input data via at least one of a input source 105, a network 110, a computer system 115, and input and output devices 118 (e.g., a monitor, a keyboard, a mouse, a printer, and the like). The network 110 may transmit video data recorded by the video input 105 to the computer system 115. Illustratively, the computer system 115 includes a CPU 120, storage 125 (e.g., a disk drive, optical disk drive, floppy disk drive, and the like), and a memory 130 containing both a computer vision engine 135 and a machine learning engine 140. As described in greater detail below, the computer vision engine 135 and the machine learning engine 140 may provide software applications configured to analyze a sequence of video frames provided by the video input 105.; [0031] Network 110 receives a source of unstructured digital video data (e.g., video stream(s), video images, or the like) from the sensor data stream video input source 105. The video input source 105 may be a video camera, a VCR, DVR, DVD, computer, an internet-based data feed web-cam device, or the like.; [0032] The computer vision engine 135 may be configured to analyze this raw information to identify active objects in the video stream, classify the objects, a source of structured digital content derive a variety of metadata regarding the actions and interactions of such objects, and supply this information to a machine learning engine 140.), and
perform representation learning by: performing self-supervised clustering of features extracted from the input data into a hierarchical set of symbols, using a neuro-symbolic generator, to generate a learned feature language ([0040] Alternatively, the estimator/identifier component 215 may derive a variety of micro features characterizing different aspects of a foreground object, e.g., size, height, width, and area (in pixels), reflectivity, shininess rigidity, speed velocity, etc. In the latter case, the machine-learning engine 140 may be configured to classify different foreground objects as being instances of a common object type, based on the similarity of one objects' micro features to others. This approach allows distinct object types to emerge from the perform representation learning by: performing self-supervised clustering of features extracted from the input data clustering of micro features (e.g., using an ART network to cluster the micro features). For example, the micro features of multiple vehicles may all be clustered as being instances of a common agent type. In such a case, the estimator/identifier component 215 does not classify an observed vehicle as being a “vehicle” directly, but instead, as being an instance of an arbitrary object type having micro features similar to other vehicles observed by the computer vision engine 135.; [0024] Thus, the semantic labeler takes the numerical data describing a foreground object and transforms it into a hierarchical set of symbols into a symbolic representation; namely, a sequence of ART network labels—each modeling a different dimension of the object.; [0043] The computer vision engine 135 may take the outputs of the components 205, 210, 215, and 220 describing the motions and actions of the tracked objects in the scene and supply this information to the machine learning engine 140. In one embodiment, the primitive event detector 212 may be configured to receive the output of the computer vision engine 135 (i.e., the video images, the object classifications, and context event stream) and generate a sequence of primitive events—labeling the observed actions or behaviors in the video with semantic meaning. For example, assume the computer vision engine 135 has identified a foreground object and classified that foreground object as being a vehicle and the context processor component 220 estimates kinematic data regarding the car's position and velocity. In such a case, this information is supplied to the machine learning engine 140 and the primitive event detector 212. In turn, the primitive event detector 212 may using a neuro-symbolic generator, to generate a learned feature language generate a semantic symbol stream providing a simple linguistic description of actions engaged in by the vehicle.); and
encoding the input data into at least one latent vector representing at least one pattern in the input data, the at least one latent vector included in the learned feature language ([0044] Illustratively, the machine learning engine 140 includes a long-term memory 225, a perceptual memory 230, an episodic memory 235, a workspace 240, codelets 245, a mapper component 211, mapper event data 255, a semantic labeler 260, and an analyzer 265. In one embodiment, the perceptual memory 230, the episodic memory 235, and the long-term memory 225 are representing at least one pattern in the input data used to identify patterns of behavior, evaluate events that transpire in the scene, and encoding the input data into at least one latent vector encode and store observations. Generally, the perceptual memory 230 receives the output of the computer vision engine 135 (e.g., the context event stream) and a primitive event stream generated by primitive event detector 212. The episodic memory 235 the at least one latent vector included in the learned feature language stores data representing observed events with details related to a particular event. That is, the episodic memory 235 may encode specific details of a particular event, i.e., “what, when, and where” something occurred within a scene, such as a particular vehicle (car A) moved to a location believed to be a parking space (parking space 5) at 9:43 AM.); and
store a representation of at least one of the at least one latent vector, the spatial anomalies, the temporal anomalies, the composite anomalies, or the at least one pattern in at least one of a semantic memory, an episodic memory, or a long- term memory ([0044] Illustratively, the store a representation of at least machine learning engine 140 includes a long-term memory 225, a perceptual memory 230, an episodic memory 235, a workspace 240, one of the at least one latent vector, the spatial anomalies, the temporal anomalies, the composite anomalies codelets 245, a mapper component 211, mapper event data 255, a or the at least one pattern in at least one of a semantic memory semantic labeler 260, and an analyzer 265. In one embodiment, the perceptual memory 230, the an episodic memory episodic memory 235, and the or a long- term memory long-term memory 225 are used to identify patterns of behavior, evaluate events that transpire in the scene, and encode and store observations.),
the system configured to (1) update the learned feature language based on newly-received input data and data retrieved from the at least one of a semantic memory, an episodic memory, or a long-term memory ([0026] In one embodiment, the machine learning engine may also include an analyzer. The semantic labeler may send a symbol trajectory to the analyzer. The symbol trajectory may be derived from observing a foreground object moving through the scene. The symbol trajectory represents semantic concepts extracted from the trajectory. Further, the analyzer may determine whether the symbol trajectory is anomalous (relative to prior observation). For example, the analyzer may compute a likelihood of observing the symbol trajectory (based on symbol trajectories previously observed in the scene). Further still, anomalous behavior of a foreground object (i.e., behavior that produces a symbol trajectory determined to be anomalous) may result in an alert passed to users of the behavioral recognition system. Consequently, the machine learning engine performs unsupervised learning to automatically determine whether a given trajectory is anomalous. This unsupervised learning and trajectory discovery are adaptive because the and data retrieved from the at least one of a semantic memory, an episodic memory, or a long-term memory knowledge about existing trajectories is the system configured to (1) update the learned feature language based on newly-received input data dynamically updated as new trajectories are observed and are classified by the machine learning engine. That is, the machine learning engine is not limited to specific pre-defined trajectory types.), and
(2) detect in real-time, and respond to, new anomalies ([0033] In general, the computer vision engine 135 and the machine learning engine 140 both detect in real-time, and respond to process video data in real-time. However, time scales for processing information by the computer vision engine 135 and the machine learning engine 140 may differ. For example, in one embodiment, the computer vision engine 135 processes the received video data frame-by-frame, while the machine learning engine 140 processes data every N-frames. In other words, while the computer vision engine 135 analyzes each frame in real-time to derive a set of information about what is occurring within a given frame, the machine learning engine 140 is not constrained by the real-time frame rate of the video input.; [0048] Once trained, the analyzer 265 may determine whether a new symbol trajectory is anomalous (relative to prior observation). If the new symbol trajectory is determined new anomalies to be anomalous, the behavioral recognition system may alert a user.).
Cobb fails to teach a decision learning engine, configured to: generate, via a plurality of codelets, a plurality of sub-decisions including detection of at least one of spatial anomalies, temporal anomalies, or composite anomalies, selectively execute codelets from the plurality of codelets in response to traversing a graph structure and satisfying associated conditions, and
Nakahata teaches a decision learning engine, configured to: generate, via a plurality of codelets, a plurality of sub-decisions including detection of at least one of spatial anomalies, temporal anomalies, or composite anomalies ([2 Spatio-temporal composition method, pg. 1029] In this method, new samples of video are broken down into small volumes that are a decision learning engine, configured to: generate, via a plurality of codelets represented by codewords from a codebook. Then, the a plurality of sub-decisions probabilities of occurrence of or composite anomalies spatio-temporal compositions formed by these codewords are calculated. Compositions with low probability are candidates to be anomalous. The training is conducted with a small sample video of a normal scene. The initial stages of sampling and creating the descriptors are identical in the training and analysis phases. Figure 1 shows the main steps of the STC method for identifying anomalies in images.; [1.1 Main contributions, pg.1028] The use of a new descriptor based on including detection of at least one of spatial anomalies, temporal anomalies both spatial and temporal gradients (see Sect. 3.2)),
selectively execute codelets from the plurality of codelets in response to traversing a graph structure and satisfying associated conditions ([2.5 Anomalous pattern detection, pg. 1034] For each vi a selectively execute codelets from the plurality of codelets codeword ck ∈ C is allocated with a similarity wi,j . The probability P(c, c , δ | EV i ) of each volume to be an anomaly is calculated in response to traversing a graph structure and satisfying associated conditions based on the spatio-temporal arrangement of the volumes within the ensemble EV i , centered in vi . For each volume the probability of occurrence is computed using Eq. (7). Volumes with a probability smaller than a given threshold, obtained experimentally, are considered to be anomalous. Figure 5 illustrates the use of the threshold. Ideally, only the anomalous points have a probability below the threshold.), and
Cobb and Nakahata are considered to be analogous to the claimed invention because they are in the same field of machine learning. In view of the teachings of Cobb, it would have been obvious for a person of ordinary skill in the art to apply the teachings of Nakahata to Cobb before the effective filing date of the claimed invention in order to allow a more reliable anomaly detection, in the identification of anomalies without the need of background subtraction, motion estimation or tracking (cf. Nakahata, [Abstract, pg. 1026] In this work, we propose improvements to the present STC method that will alleviate this problem in two ways. First, a two-stage dictionary learning process is performed to allow a more reliable anomaly detection. Second, improved spatio-temporal features are employed. These modified features are extracted after an enhanced temporal filtering that performs a temporal regularization of the video sequence. The proposed approach gives very good results in the identification of anomalies without the need of background subtraction, motion estimation or tracking. The results are shown to be comparable or even superior to those of other state-of-the-art methods using bag-of-video words or other moving-camera surveillance methods. The system is accurate even with no prior knowledge of the type of event to be observed, being robust to cluttered environments, as illustrated by several practical examples. These results are obtained without compromising the performance of the algorithm in the static cameras case.).
Regarding claim 21, Cobb, as modified by Nakahata, teaches The system of claim 8.
Nakahata teaches wherein the neuro-symbolic generator is configured to generate at least one of a-symbols, p-symbols, or y-symbols of variable size ([2.3 Codebook, pg. 1030-1031] After creating the codebook, neuro-symbolic generator is configured to generate at least one of a-symbols each volume vi is related to a codeword c j with a weight wi,j given by where d(vi, c j) is the Euclidean distance between the volume vi and the codeword c j.; [2.4 Spatio-temporal composition, pg. 1031-1032] Thus, the ensemble of volumes Ei , centered at position (xi, yi, ti), is initially represented as a set of video volumes and their relative positions with respect to the central volume: Ei = {ΔEi vk , vk , vi}K k=1. (4) Each volume vk of the set is linked with the codeword c j ∈ Cwith a weightwj p-symbols representing their similarity. Thus, the arrangement of volumes may be represented by a set of codewords and their spatio-temporal arrangement. Let ν ⊂ Rnx×ny×nt be the spatio of descriptors of a video volume, and C the codebook; c : ν → C defines a random variable that allocates a codeword to a volume of video and c : ν → C defines a random variable designating a codeword to the volume in the center of the ensemble. In addition, δ : R3 → R3 defines a random variable representing the relative distance from the central point associated with codeword c to the point associated with codeword c. Therefore, the ensemble of volumes can be represented as an arrangement of words of the codebook, as shown in Fig. 3b. In other words, instead of representing the Ei as an arrangement of volumes, it is y-symbols of variable size represented as a codeword arrangement.).
Cobb and Nakahata are combinable for the same rationale as set forth above with respect to claim 8.
Regarding claim 22, Cobb, as modified by Nakahata, teaches The system of claim 21.
Nakahata teaches wherein the neuro-symbolic generator is configured to generate the p-symbols through the association of a-symbols based on at least one co-occurrence pattern ([2.3 Codebook, pg. 1030-1031] After creating the codebook, each volume vi is related to a codeword c j with a weight wi,j given by where d(vi, c j) is the Euclidean distance between the volume vi and the codeword c j.; [2.4 Spatio-temporal composition, pg. 1031-1032] Thus, the ensemble of volumes Ei , centered at position (xi, yi, ti), is initially represented as a set of video volumes and their relative positions with respect to the central volume: Ei = {ΔEi vk , vk , vi}K k=1. (4) Each configured to generate the p-symbols through the association of a-symbols based on at least one co-occurrence pattern volume vk of the set is linked with the codeword c j ∈ Cwith a weightwj representing their similarity. Thus, the arrangement of volumes may be represented by a set of codewords and their spatio-temporal arrangement. Let ν ⊂ Rnx×ny×nt be the spatio of descriptors of a video volume, and C the codebook; c : ν → C defines a random variable that allocates a codeword to a volume of video and c : ν → C defines a random variable designating a codeword to the volume in the center of the ensemble. In addition, δ : R3 → R3 defines a random variable representing the relative distance from the central point associated with codeword cto the point associated with codeword c. Therefore, the ensemble of volumes can be represented as an arrangement of words of the codebook, as shown in Fig. 3b. In other words, instead of representing the Ei as an arrangement of volumes, it is represented as a codeword arrangement.).
Cobb and Nakahata are combinable for the same rationale as set forth above with respect to claim 8.
Regarding claim 23, Cobb, as modified by Nakahata, teaches The system of claim 21.
Nakahata teaches wherein the neuro-symbolic generator is configured to generate the y-symbols through the aggregation of p-symbols into a representation of at least one statistically significant feature relationship ([2.3 Codebook, pg. 1030-1031] After creating the codebook, each volume vi is related to a codeword c j with a weight wi,j given by where d(vi, c j) is the Euclidean distance between the volume vi and the codeword c j.; [2.4 Spatio-temporal composition, pg. 1031-1032] Thus, the ensemble of volumes Ei , centered at position (xi, yi, ti), is initially represented as a set of video volumes and their relative positions with respect to the central volume: Ei = {ΔEi vk , vk , vi}K k=1. (4) Each volume vk of the set is linked with the codeword c j ∈ Cwith a weightwj representing their similarity. Thus, the arrangement of volumes may be represented by a set of codewords and their spatio-temporal arrangement. Let ν ⊂ Rnx×ny×nt be the spatio of descriptors of a video volume, and C the codebook; c : ν → C defines a random variable that allocates a codeword to a volume of video and c : ν → C defines a random variable designating a codeword to the volume in the center of the ensemble. In addition, δ : R3 → R3 defines a random variable representing the relative distance from the central point associated with codeword c to the point associated with codeword c. Therefore, the configured to generate the y-symbols through the aggregation of p-symbols into a representation of at least one statistically significant feature relationship ensemble of volumes can be represented as an arrangement of words of the codebook, as shown in Fig. 3b. In other words, instead of representing the Ei as an arrangement of volumes, it is represented as a codeword arrangement.).
Cobb and Nakahata are combinable for the same rationale as set forth above with respect to claim 8.
Regarding claim 24, Cobb, as modified by Nakahata, teaches The system of claim 8.
Nakahata teaches wherein the plurality of sub-decisions further includes at least one of classification, prediction, or condition generation, based on the representations of the at least one of the at least one latent vector, the spatial anomalies, the temporal anomalies, the composite anomalies, or the at least one pattern ([2.5 Anomalous pattern detection, pg. 1033-1034] In the analysis phase the steps of sampling and descriptor calculation are the same as in the training phase. Then, using the codebook created in the training phase, the distance between the volume and every codeword is computed using Eq. (2). based on the representations of the at least one of the at least one latent vector, the spatial anomalies, the temporal anomalies, the composite anomalies, or the at least one pattern Equation (6) is the codeword probability assignment, that is dependent on the relations between the central volume and the other volumes vk in the ensemble Ei . Given a video of interest V, EV i is an ensemble of video volumes centered at point (xi, yi, ti) and vi is the central volume of this ensemble. The probability of the volume vi can be written as: P c, c , δ | EV i = K k P δ | c, c , ΔEV i vk P (c | vk ) P c | vi , (7) where vk is a volume inside EV i , ΔEV i vk is the relative position of the volume vk , c is the codeword attributed to vi , c is the codeword attributed to vk and δ is the relative distance of the codeword in the codeword space. The term P(δ | c, c , ΔEV i vk ) is the probability of the spatio-temporal arrangement, whose pdf is calculated as given in Sect. 2.4. The a posteriori probability is calculated according to P c j | vi = wi,j × P(c j) j wi,j × P(c j) , (8) where the weight w is given by Eq. (2). In brief, the video V to be analyzed is densely sampled into video volumes vi . For each vi a codeword ck ∈ C is allocated with a similarity wi,j . The plurality of sub-decisions further includes at least one of prediction probability P(c, c , δ | EV i ) of each volume classification to be an anomaly is calculated based on the spatio-temporal condition generation arrangement of the volumes within the ensemble EV i , centered in vi . For each volume the probability of occurrence is computed using Eq. (7). Volumes with a probability smaller than a given threshold, obtained experimentally, are considered to be anomalous. Figure 5 illustrates the use of the threshold. Ideally, only the anomalous points have a probability below the threshold.).
Cobb and Nakahata are combinable for the same rationale as set forth above with respect to claim 8.
Claims 9, 16, 19 are rejected under 35 U.S.C. 103 as being unpatentable over Cobb, in view of Nakahata, and further in view of Knight et al. (U.S. Pre-Grant Publication No. 20050171948, hereinafter ‘Knight').
Regarding claim 9 and analogous claim 16, Cobb, as modified by Nakahata, teaches The system of claim 8.
Cobb, as modified by Nakahata, fails to teach wherein the decision learning engine is configured to convert unstructured data into structured data.
Knight teaches wherein the decision learning engine is configured to convert unstructured data into structured data ([0052] The individual data collections 17, 20; 26, 29 each constitute a semantically- related collection of stored data, including all forms and types of unstructured and semi-structured (textual) data, including electronic message stores, such as electronic mail (email) folders, word processing documents or Hypertext documents, and could also include graphical or multimedia data. The unstructured data also includes genome and protein sequences and similar data collections. The data collections include some form of vocabulary with which atomic data units are defined and features are semantically-related by a grammar, as would be recognized by one skilled in the art. An atomic data unit is analogous to a feature and consists of one or more searchable characteristics which, when taken singly or in combination, represent a grouping having a common semantic meaning. The grammar allows the features to be combined syntactically and semantically and enables the discovery of latent semantic meanings. The documents could also be in the decision learning engine is configured to convert unstructured data into structured data form of structured data, such as stored in a spreadsheet or database. Content mined from these types of documents will not require preprocessing, as described below.).
Cobb, Nakahata, and Knight are considered to be analogous to the claimed invention because they are in the same field of machine learning. In view of the teachings of Cobb and Nakahata, it would have been obvious for a person of ordinary skill in the art to apply the teachings of Knight to Cobb before the effective filing date of the claimed invention in order to provide an approach for ordered set of extracted features determined from a multi-dimensional problem space and transforming feature space into ordered scale space to provide scalable feature space capable of abstraction (cf. Knight, [0011] Therefore, there is a need for an approach to providing an ordered set of extracted features determined from a multi-dimensional problem space, including text documents and genome and protein sequences. Preferably, such an approach will isolate critical feature spaces while filtering out null valued, conceptually insignificant, and redundant features within the concept space.; [0012] There is a further need for an approach that transforms the feature space into an ordered scale space. Preferably, such an approach would provide a scalable feature space capable of abstraction in varying levels of detail through multiresolution analysis.).
Regarding claim 19, Cobb, as modified by Nakahata, teaches The non-transitory, processor-readable medium of claim 15.
Cobb, as modified by Nakahata, fails to teach wherein the representation learning output includes a learned pattern.
Knight teaches wherein the representation learning output includes a learned pattern ([0073] FIG. 9 is a flow diagram showing the routine 120 for determining a frequency of concepts for use in the routine of FIG. 8. The purpose of this routine is to extract individual features from each data collection and to create a normalized representation of the feature occurrences and co-occurrences on a per-data collection basis.; [0013] The present invention provides a system and method for transforming a multi-dimensional feature space into an ordered and prioritized scale space representation. The scale space will generally be defined in Hilbert function space. A multiplicity of individual features are extracted from a plurality of discrete data collections. Each individual feature represents latent content inherent in the semantic structuring of the data collection.; [0014] An embodiment provides a system and method for identifying critical features in an ordered scale space within a multi-dimensional feature space. Features are extracted from a plurality of data collections. Each data collection is characterized by a collection of features semantically-related by a grammar. Each feature is then normalized and frequencies of occurrence and co-occurrences for the features for each of the data collections is determined. The occurrence frequencies and the co-occurrence frequencies for each of the extracted features are representation learning output includes a learned pattern mapped into a set of patterns of occurrence frequencies and a set of patterns of co-occurrence frequencies. The pattern for each data collection is selected and similarity measures between each occurrence frequency in the selected pattern is calculated.).
Cobb, Nakahata, and Knight are combinable for the same rationale as set forth above with respect to claim 9.
Claims 12, 18 are rejected under 35 U.S.C. 103 as being unpatentable over Cobb, in view of Nakahata, and further in view of Seow et al. (U.S. Pre-Grant Publication No. 20170293609, hereinafter ‘Seow2').
Regarding claim 12 and analogous claim 18, Cobb, as modified by Nakahata, teaches The system of claim 8.
Cobb, as modified by Nakahata, fails to teach wherein the input data includes the sensor data stream, and the sensor data stream includes data from at least one Internet of Things (IoT) sensor.
Seow2 teaches wherein the input data includes the sensor data stream, and the sensor data stream includes data from at least one Internet of Things (IoT) sensor ([0032] FIG. 1 illustrates components of a neuro-linguistic cognitive AI system 100, according to an embodiment. As shown, the Cognitive AI System 100 includes one or more input data includes the sensor data stream input source devices 105 (e.g., sensor devices), a network 110, and one or more computer systems 115. The network 110 may transmit data input by the input source devices 105 to the computer system 115.; [0080] Advantageously, techniques disclosed herein may be used to monitor observations from input source devices, for example, sensors such as video surveillance systems, SCADA systems, data network security systems, the sensor data stream includes data from at least one Internet of Things (IoT) sensor internet of things (TOT), and the like, and generate special event notifications of anomalous observations.).
Cobb, Nakahata, and Seow2 are considered to be analogous to the claimed invention because they are in the same field of machine learning. In view of the teachings of Cobb and Nakahata, it would have been obvious for a person of ordinary skill in the art to apply the teachings of Seow2 to Cobb before the effective filing date of the claimed invention in order to monitor observations from input sources to generate notifications of anomalous observations (cf. Seow2, [0080] Advantageously, techniques disclosed herein may be used to monitor observations from input source devices, for example, sensors such as video surveillance systems, SCADA systems, data network security systems, internet of things (TOT), and the like, and generate special event notifications of anomalous observations.).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Xu et al. (U.S. Pre-Grant Publication No. 20190095417) teaches a computer-implemented method, system, and computer program product provided for content aware heterogeneous log pattern comparative analysis.
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 extension fee 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 date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MAGGIE MAIDO whose telephone number is (703) 756-1953. The examiner can normally be reached M-Th: 6am - 4pm.
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, Michael Huntley can be reached on (303) 297-4307. 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.
/MM/Examiner, Art Unit 2129
/MICHAEL J HUNTLEY/Supervisory Patent Examiner, Art Unit 2129