International Business Machines Corporation

Toshiba Global Commerce Solutions Holdings Corporation

Massachusetts Institute of Technology

Globalfoundries Inc.

Toshiba Global Commerce

Solutions Holdings Corporation

Quanfu Fan

Sharathchandra UmapathiRao Pankanti

Sachiko Miyazawa

Yun Zhai

Russell Patrick Bobbitt

Arun Hampapur

Akira Yanagawa

Richard Chen

Frederik C Kjeldsen

Lei Ding

Hoang Trinh

Zuoxuan Lu

Sijia Liu

Sharathchandra Umapathirao Pankanti

Rameswar Panda

Prasad Gabbur

Rogerio Schmidt Feris

Chuang Gan

Lisa Maire Brown

Dan Gutfreund

Sharathchandra U Pankanti, I

Jingjing Liu

Jiyan Pan

Marco Pistoia

Dakuo Wang

Toyotaro Suzumura

Chung-Ching Lin

Mandis S Beigi

Frederik C M Kjeldsen

Gaoyuan Zhang

Honghui Shi

Hildegard Kuehne

Tete Xiao

Bolei Zhou

Abhijit S Mudigonda

Zuoxuan Max Lu

One embodiment of the invention provides a method for video recognition. The method comprises receiving an input video comprising a sequence of video segments over a plurality of data modalities. The method further comprises, for a video segment of the sequence, selecting one or more data modalities based on data representing the video segment. Each data modality selected is optimal for video recognition of the video segment. The method further comprises, for each data modality selected, providing at least one data input representing the video segment over the data modality selected to a machine learning model corresponding to the data modality selected, and generating a first type of prediction representative of the video segment via the machine learning model. The method further comprises determining a second type of prediction representative of the entire input video by aggregating all first type of predictions generated.

Adaptive selection of data modalities for efficient video recognition

A system and a method for tracking a plurality of objects, including obtaining input data, estimating a number of skipping frames of the input data based on information from the input data, predicting results based on the estimating of the number of skipping frames, and correcting the predicted results.

System and method for multiple object tracking

Techniques and apparatus for analyzing visual content using a visual transformer are described. An example technique includes generating a first set of tokens based on a visual content item. Each token in the first set of tokens is associated with a regional feature from a different region of a plurality of regions of the visual content item. A second set of tokens is generated based on the visual content item. Each token in the second set of tokens is associated with a local feature from one of the plurality of regions of the visual content item. At least one feature map is generated for the visual content item, based on analyzing the first set of tokens and the second set of tokens separately using a hierarchical vision transformer. At least one vision task is performed based on the at least one feature map.

Regional-to-local attention for vision transformers

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.

Detecting hybdrid-distance adversarial patches

Techniques are described for reducing the number of parameters of a deep neural network model. According to one or more embodiments, a device can comprise a memory that stores computer executable components and a processor that executes the computer executable components stored in the memory. The computer executable components can comprise a structure extraction component that determines a number of input nodes associated with a fully connected layer of a deep neural network model. The computer executable components can further comprise a transformation component that replaces the fully connected layer with a number of sparsely connected sublayers, wherein the sparsely connected sublayers have fewer connections than the fully connecter layer, and wherein the number of sparsely connected sublayers is determined based on a defined decrease to the number of input nodes.

Restructuring deep neural networks to reduce the number of parameters

Embodiments of the present invention are directed to a computer-implemented method for multiscale representation of input data. A non-limiting example of the computer-implemented method includes a processor receiving an original input. The processor downsamples the original input into a downscaled input. The processor runs a first convolutional neural network ('CNN') on the downscaled input. The processor runs a second CNN on the original input, where the second CNN has fewer layers than the first CNN. The processor merges the output of the first CNN with the output of the second CNN and provides a result following the merging of the outputs.

Multiscale feature representations for object recognition and detection

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.

Root cause analysis of vulnerability of neural networks to adversarial examples

Systems and methods for performing video understanding and analysis. Sets of feature maps for high resolution images and low resolution images in a time sequence of images are combined into combined sets of feature maps each having N feature maps. A time sequence of temporally aggregated sets of feature maps is created for each combined set of feature maps by: selecting a selected combined set of feature maps corresponding to an image at time 't' in the time sequence of images; applying, by channel-wise multiplication, a feature map weighting vector to a number of combined sets of feature maps that are temporally adjacent to the selected combined set of feature maps; and summing elements of the number of combined set of feature maps into a temporally aggregated set of feature maps. The time sequence of temporally aggregated sets of feature maps is processed to perform video understanding processing.

Systems and approaches for learning efficient representations for video understanding

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.

Quantifying vulnerabilities of deep learning computing systems to adversarial perturbations

Provided are embodiments including a computer-implemented method for performing recognition. The computer-implemented method includes receiving video data, and performing, at a pre-attention prediction module, a pre-attention prediction for the video data to generate first prediction priors. The computer-implemented method also includes receiving, at a dual attention module, data including the video data and data from the pre-attention prediction to generate attention maps, wherein the attention maps indicate a region of interest of a frame of the video data, wherein the dual attention module generates enhanced feature representations, and performing, at a post-attention prediction module, a post-attention prediction from data from the dual attention module based at least in part on the enhanced feature representation. Also provided are embodiments for a system and a computer program produce for performing recognition.

Context-aware action recognition by dual attention networks

Technologies for providing convolutional neural networks are described. An analysis component determines an initial convolutional layer in a network architecture of a convolutional neural network and one or more subsequent convolutional layers in the network architecture. A replacement component replaces original convolutional kernels in the initial convolutional layer with initial sparse convolutional kernels, and replaces subsequent convolutional kernels in one or more subsequent convolutional layers with complementary sparse convolutional kernels. The complementary sparse kernels have a complementary pattern with respect to sparse kernels of a previous convolutional layer. Analyzing the network architecture and a trained model of a convolutional neural network can determine the original convolutional kernels and replace those kernels with sparse kernels based on similarity and/or weight in an initial layer, with sparse complementary kernels used in subsequent layers.

Convolutional neural network with sparse and complementary kernels

An image data is convolved with one or more kernels and corresponding one or more feature maps generated. Region of interest maps are extracted from the one or more feature maps, and pooled based on one or more features selected as selective features. Pooling generates a feature vector with dimensionality less than a dimensionality associated with the one or more feature maps. The feature vector is flattened and input as a layer in a neural network. The neural network outputs a classification associated with an object in the image data.

Sparse region-of-interest pooling for object detection

Techniques facilitating generation of a fused kernel that can approximate a full kernel of a convolutional neural network are provided. In one example, a computer-implemented method comprises determining a first pattern of samples of a first sample matrix and a second pattern of samples of a second sample matrix. The first sample matrix can be representative of a sparse kernel, and the second sample matrix can be representative of a complementary kernel. The first pattern and second pattern can be complementary to one another. The computer-implemented method also comprises generating a fused kernel based on a combination of features of the sparse kernel and features of the complementary kernel that are combined according to a fusing approach and training the fused kernel.

Fusing sparse kernels to approximate a full kernel of a convolutional neural network

Embodiments of the present invention provide a system, method, and program product to determine whether a product has been successfully purchased by identifying in a video record when a movement of a product adjacent to a scanner occurs, and whether the scanner did not record a purchase transaction at that time; measuring a difference in time between the time of the movement of the product and a time of another movement of a product, and determining by a trained support vector machine a likelihood that the product was successfully purchased. Alternately, the difference in time can be measured between the time of the movement of the product and a time of a transaction record, or between the time of the movement of the product and a boundary time. The support vector machine can use a radial basis function kernel and can generate a decision value and a confidence score.

Systems and methods for false alarm reduction during event detection

A method to automatically calibrate a static camera in a vehicle is provided. The method may include receiving a captured image file. The method may also include detecting a plurality of vehicles in the captured image file. The method may further include determining a 2D projected shape, size, location, direction of travel, and a plurality of features for each vehicle at various locations in the captured image file. The method may additionally include inferring a plurality of geometry scenes associated with the captured image file, whereby the plurality of geometry scenes is inferred based on the determined 2D projected shape, size, location, direction of travel, and the plurality of features of each vehicle within the detected plurality of vehicles as projected onto the captured image file. The method may include calibrating the static camera based on the inferred plurality of geometry scenes associated with the captured image file.

Growing community of inventors

Somerville, MA, United States of America

Quanfu Fan