-
公开(公告)号：US20220067505A1
公开(公告)日：2022-03-03
申请号：US17005144
申请日：2020-08-27
Inventor： Ao Liu , Sijia Liu , Bo Wu , Lirong Xia , Qi Cheng Li , Chuang Gan
Abstract: Interpretation maps of convolutional neural networks having certifiable robustness using Rényi differential privacy are provided. In one aspect, a method for generating an interpretation map includes: adding generalized Gaussian noise to an image x to obtain T noisy images, wherein the generalized Gaussian noise constitutes perturbations to the image x; providing the T noisy images as input to a convolutional neural network; calculating T noisy interpretations of output from the convolutional neural network corresponding to the T noisy images; re-scaling the T noisy interpretations using a scoring vector ν to obtain T re-scaled noisy interpretations; and generating the interpretation map using the T re-scaled noisy interpretations, wherein the interpretation map is robust against the perturbations.
-
12.发明申请公开(公告)号：US20200285952A1
公开(公告)日：2020-09-10
申请号：US16296897
申请日：2019-03-08
Applicant: International Business Machines Corporation
Inventor： Sijia Liu , Quanfu Fan , Chuang Gan , Dakuo Wang
Abstract: Mechanisms are provided for generating an adversarial perturbation attack sensitivity (APAS) visualization. The mechanisms receive a natural input dataset and a corresponding adversarial attack input dataset, where the adversarial attack input dataset comprises perturbations intended to cause a misclassification by a computer model. The mechanisms determine a sensitivity measure of the computer model to the perturbations in the adversarial attack input dataset based on a processing of the natural input dataset and corresponding adversarial attack input dataset by the computer model. The mechanisms generate a classification activation map (CAM) for the computer model based on results of the processing and a sensitivity overlay based on the sensitivity measure. The sensitivity overlay graphically represents different classifications of perturbation sensitivities. The mechanisms apply the sensitivity overlay to the CAM to generate and output a graphical visualization output of the computer model sensitivity to perturbations of adversarial attacks.
-
公开(公告)号：US12242980B2
公开(公告)日：2025-03-04
申请号：US17015243
申请日：2020-09-09
Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION
Inventor： Parikshit Ram , Dakuo Wang , Deepak Vijaykeerthy , Vaibhav Saxena , Sijia Liu , Arunima Chaudhary , Gregory Bramble , Horst Cornelius Samulowitz , Alexander Gray
IPC: G06N5/04 , G06F9/38 , G06F18/243
Abstract: The exemplary embodiments disclose a method, a computer program product, and a computer system for determining that one or more model pipelines satisfy one or more constraints. The exemplary embodiments may include detecting a user uploading data, one or more constraints, and one or more model pipelines, collecting the data, the one or more constraints, and the one or more model pipelines, and determining that one or more of the model pipelines satisfies all of the one or more constraints based on applying one or more algorithms to the collected data, constraints, and model pipelines.
-
公开(公告)号：US11836256B2
公开(公告)日：2023-12-05
申请号：US16256107
申请日：2019-01-24
Applicant: International Business Machines Corporation
Inventor： Pin-Yu Chen , Sijia Liu , Lingfei Wu , Chia-Yu Chen
IPC: G06F21/57 , G06N3/04 , G06N3/08 , G06V10/764 , G06V10/82
CPC classification number: G06F21/577 , G06N3/04 , G06N3/08 , G06V10/764 , G06V10/82 , G06F2221/034
Abstract: An adversarial robustness testing method, system, and computer program product include testing a robustness of a black-box system under different access settings via an accelerator.
-
公开(公告)号：US11687777B2
公开(公告)日：2023-06-27
申请号：US17005144
申请日：2020-08-27
Inventor： Ao Liu , Sijia Liu , Bo Wu , Lirong Xia , Qi Cheng Li , Chuang Gan
CPC classification number: G06N3/08 , G06F16/56 , G06F18/21 , G06T3/4046 , G06T5/002 , G06T2207/20084
Abstract: Interpretation maps of convolutional neural networks having certifiable robustness using Rényi differential privacy are provided. In one aspect, a method for generating an interpretation map includes: adding generalized Gaussian noise to an image x to obtain T noisy images, wherein the generalized Gaussian noise constitutes perturbations to the image x; providing the T noisy images as input to a convolutional neural network; calculating T noisy interpretations of output from the convolutional neural network corresponding to the T noisy images; re-scaling the T noisy interpretations using a scoring vector υ to obtain T re-scaled noisy interpretations; and generating the interpretation map using the T re-scaled noisy interpretations, wherein the interpretation map is robust against the perturbations.
-
Patent Agency Ranking
16.发明授权公开(公告)号：US11625487B2
公开(公告)日：2023-04-11
申请号：US16256267
申请日：2019-01-24
Inventor： Pin-Yu Chen , Sijia Liu , Akhilan Boopathy , Tsui-Wei Weng , Luca Daniel
Abstract: A certification method, system, and computer program product include certifying an adversarial robustness of a convolutional neural network by deriving an analytic solution for a neural network output using an efficient upper bound and an efficient lower bound on an activation function and applying the analytic solution in computing a certified robustness.
-
公开(公告)号：US20230005111A1
公开(公告)日：2023-01-05
申请号：US17363054
申请日：2021-06-30
Applicant: International Business Machines Corporation
Inventor： Quanfu Fan , Sijia Liu , Richard Chen , Rameswar Panda
Abstract: A hybrid-distance adversarial patch generator can be trained to generate a hybrid adversarial patch effective at multiple distances. The hybrid patch can be inserted into multiple sample images, each depicting an object, to simulate inclusion of the hybrid patch at multiple distances. The multiple sample images can then be used to train an object detection model to detect the objects.
-
公开(公告)号：US11443069B2
公开(公告)日：2022-09-13
申请号：US16559161
申请日：2019-09-03
Applicant: International Business Machines Corporation
Inventor： Sijia Liu , Quanfu Fan , Gaoyuan Zhang , Chuang Gan
Abstract: An illustrative embodiment includes a method for protecting a machine learning model. The method includes: determining concept-level interpretability of respective units within the model; determining sensitivity of the respective units within the model to an adversarial attack; identifying units within the model which are both interpretable and sensitive to the adversarial attack; and enhancing defense against the adversarial attack by masking at least a portion of the units identified as both interpretable and sensitive to the adversarial attack.
-
公开(公告)号：US20200286243A1
公开(公告)日：2020-09-10
申请号：US16292847
申请日：2019-03-05
Applicant: International Business Machines Corporation
Inventor： Chuang Gan , Yang Zhang , Sijia Liu , Dakuo Wang
Abstract: Embodiments of the present invention are directed to a computer-implemented method for action localization. A non-limiting example of the computer-implemented method includes receiving, by a processor, a video and segmenting, by the processor, the video into a set of video segments. The computer-implemented method classifies, by the processor, each video segment into a class and calculates, by the processor, importance scores for each video segment of a class within the set of video segments. The computer-implemented method determines, by the processor, a winning video segment of the class within the set of video segments based on the importance scores for each video segment within the class, stores, by the processor, the winning video segment from the set of video segments, and removes the winning video segment from the set of video segments.
-
公开(公告)号：US20200175281A1
公开(公告)日：2020-06-04
申请号：US16206683
申请日：2018-11-30
Applicant: International Business Machines Corporation
Inventor： Chuang Gan , Sijia Liu , Dakuo Wang , Yang Zhang
Abstract: A method (and structure and computer product) of temporal action localization in video data includes receiving a stream of video data and determining all proposals in the video data stream, the proposals being candidate regions for temporal action in the video data stream. Values for a pair-wise relation function are calculated for relating the proposals, wherein the pair-wise relation function calculates a scalar value representing a pair-wise relation weight for pairs of the proposals.
-
-
-
-
-
-
-
-
-
Search Results
40
records
(took 0.023 seconds)
