Nec Corporation

At+t Corp.

AT&T Intellectual Property I, LP

Nec Laboratories America, Inc.

Lucent Technologies Inc.

Ncr Corporation

Nec C Orporation

Henry Spalding Baird

John Stewart Denker

Donnie Henderson

Hans Peter Graf

Wayne E. Hubbard

Richard Edwin Howard

Lawrence David Jackel

Lawrence O'gorman

Eric Cosatto

Renqiang Min

Joern Ostermann

Asim Kadav

Thomas M Isaacson

Igor Durdanovic

Juergen Schroeter

Shan Liu

David Crawford Gibbon

Andrea Basso

Farley Lai

Iain Melvin

Mehmet Reha Civanlar

Volker Franz Strom

Yann Andre LeCun

Christopher Malon

Fu Jie Huang

Alexandru Niculescu-Mizil

Gerasimos Potamianos

Kai Li

Ligong Han

Srimat T Chakradhar

Srihari Cadambi

Bing Bai

Barbara Buda

Vladimir Vapnik

Hao Li

Xavier Fontaine

Mehmat Reha Civanlar

Wayne E Hubbard

Yi Huang

Eric D Petajan

Erik Kruus

Daniel Jitzchak Mayer

Christopher John Burges

Venkata S Jakkula

Murugan Sankaradass

Michela Becchi

Michael Alexander Snower

Abhinandan Majumdar

Leon Bottou

Meera Hahn

Yizhe Zhu

Volker Franz Storm

Honglu Zhou

Chih-Yao Ma

Yiren Jian

A method generates new binding peptides to Major Histocompatibility Complex (MHC) proteins. The method includes training a Generative Adversarial Network (GAN) having a generator and a discriminator only on a set of binding peptide sequences given training data comprising the set of binding peptide sequences and a set of non-binding peptide sequences. A GAN training objective includes the discriminator being iteratively updated to distinguish generated peptide sequences from sampled binding peptide sequences as fake or real and the generator being updated to fool the discriminator. The GAN training objective is optimized while learning two projection vectors for a binding class with two cross-entropy losses. A first loss discriminates binding peptide sequences in the training data from non-binding peptide sequences in the training data. A second loss discriminates generated binding peptide sequences from non-binding peptide sequences in the training data. The method enables supporting decision-making related to new binding peptide generation.

Peptide based vaccine generation system with dual projection generative adversarial networks

A method generates new binding peptides to Major Histocompatibility Complex (MHC) proteins. The method includes training, a Generative Adversarial Network (GAN) having a generator and a discriminator only on a set of binding peptide sequences given training data comprising the set of binding peptide sequences and a set of non-binding peptide sequences. A GAN training objective includes the discriminator being iteratively updated to distinguish generated peptide sequences from sampled binding peptide sequences as fake or real and the generator being updated to fool the discriminator. The GAN training objective is optimized while learning two projection vectors for a binding class with two cross-entropy losses. A first loss discriminates binding peptide sequences in the training data from non-binding peptide sequences in the training data. A second loss discriminates generated binding peptide sequences from non-binding peptide sequences in the training data. The method enables supporting decision-making related to new binding peptide generation.

A method for implementing source-free domain adaptive detection is presented. The method includes, in a pretraining phase, applying strong data augmentation to labeled source images to produce perturbed labeled source images and training an object detection model by using the perturbed labeled source images to generate a source-only model. The method further includes, in an adaptation phase, training a self-trained mean teacher model by generating a weakly augmented image and multiple strongly augmented images from unlabeled target images, generating a plurality of region proposals from the weakly augmented image, selecting a region proposal from the plurality of region proposals as a pseudo ground truth, detecting, by the self-trained mean teacher model, object boxes and selecting pseudo ground truth boxes by employing a confidence constraint and a consistency constraint, and training a student model by using one of the multiple strongly augmented images jointly with an object detection loss.

Source-free cross domain detection method with strong data augmentation and self-trained mean teacher modeling

Systems and methods for predicting T-Cell receptor (TCR)-peptide interaction, including training a deep learning model for the prediction of TCR-peptide interaction by determining a multiple sequence alignment (MSA) for TCR-peptide pair sequences from a dataset of TCR-peptide pair sequences using a sequence analyzer, building TCR structures and peptide structures using the MSA and corresponding structures from a Protein Data Bank (PDB) using a MODELLER, and generating an extended TCR-peptide training dataset based on docking energy scores determined by docking peptides to TCRs using physical modeling based on the TCR structures and peptide structures built using the MODELLER. TCR-peptide pairs are classified and labeled as positive or negative pairs using pseudo-labels based on the docking energy scores, and the deep learning model is iteratively retrained based on the extended TCR-peptide training dataset and the pseudo-labels until convergence.

T-cell receptor repertoire selection prediction with physical model augmented pseudo-labeling for personalized medicine decision making

A method for learning disentangled representations of videos is presented. The method includes feeding each frame of video data into an encoder to produce a sequence of visual features, passing the sequence of visual features through a deep convolutional network to obtain a posterior of a dynamic latent variable and a posterior of a static latent variable, sampling static and dynamic representations from the posterior of the static latent variable and the posterior of the dynamic latent variable, respectively, concatenating the static and dynamic representations to be fed into a decoder to generate reconstructed sequences, and applying three regularizers to the dynamic and static latent variables to trigger representation disentanglement. To facilitate the disentangled sequential representation learning, orthogonal factorization in generative adversarial network (GAN) latent space is leveraged to pre-train a generator as a decoder in the method.

Learning orthogonal factorization in GAN latent space

A computer-implemented method is provided for action localization. The method includes converting one or more video frames into person keypoints and object keypoints. The method further includes embedding position, timestamp, instance, and type information with the person keypoints and object keypoints to obtain keypoint embeddings. The method also includes predicting, by a hierarchical transformer encoder using the keypoint embeddings, human actions and bounding box information of when and where the human actions occur in the one or more video frames.

Keypoint based action localization

Methods and systems for disentangled data generation include accessing a dataset including pairs, each formed from a given input text structure and a given style label for the input text structures. An encoder is trained to disentangle a sequential text input into disentangled representations, including a content embedding and a style embedding, based on a subset of the dataset, using an objective function that includes a regularization term that minimizes mutual information between the content embedding and the style embedding. A generator is trained to generate a text output that includes content from the style embedding, expressed in a style other than that represented by the style embedding of the text input.

Contextual text generation for question answering and text summarization with supervised representation disentanglement and mutual information minimization

A method for using a multi-hop reasoning framework to perform multi-step compositional long-term reasoning is presented. The method includes extracting feature maps and frame-level representations from a video stream by using a convolutional neural network (CNN), performing object representation learning and detection, linking objects through time via tracking to generate object tracks and image feature tracks, feeding the object tracks and the image feature tracks to a multi-hop transformer that hops over frames in the video stream while concurrently attending to one or more of the objects in the video stream until the multi-hop transformer arrives at a correct answer, and employing video representation learning and recognition from the objects and image context to locate a target object within the video stream.

Multi-hop transformer for spatio-temporal reasoning and localization

A computer-implemented method executed by at least one processor for performing mini-batching in deep learning by improving cache utilization is presented. The method includes temporally localizing a candidate clip in a video stream based on a natural language query, encoding a state, via a state processing module, into a joint visual and linguistic representation, feeding the joint visual and linguistic representation into a policy learning module, wherein the policy learning module employs a deep learning network to selectively extract features for select frames for video-text analysis and includes a fully connected linear layer and a long short-term memory (LSTM), outputting a value function from the LSTM, generating an action policy based on the encoded state, wherein the action policy is a probabilistic distribution over a plurality of possible actions given the encoded state, and rewarding policy actions that return clips matching the natural language query.

Efficient and fine-grained video retrieval

Aspects of the present disclosure describe systems, methods and structures for an efficient multi-person posetracking method that advantageously achieves state-of-the-art performance on PoseTrack datasets by only using keypoint information in a tracking step without optical flow or convolution routines. As a consequence, our method has fewer parameters and FLOPs and achieves faster FPS. Our method benefits from our parameter-free tracking method that outperforms commonly used bounding box propagation in top-down methods. Finally, we disclose tokenization and embedding multi-person pose keypoint information in the transformer architecture that can be re-used for other pose tasks such as pose-based action recognition.

Keypoint based pose-tracking using entailment

A computer-implemented method is provided for disentangled data generation. The method includes accessing, by a variational autoencoder, a plurality of supervision signals. The method further includes accessing, by the variational autoencoder, a plurality of auxiliary tasks that utilize the supervision signals as reward signals to learn a disentangled representation. The method also includes training the variational autoencoder to disentangle a sequential data input into a time-invariant factor and a time-varying factor using a self-supervised training approach which is based on outputs of the auxiliary tasks obtained by using the supervision signals to accomplish the plurality of auxiliary tasks.

Self-supervised sequential variational autoencoder for disentangled data generation

A method is provided for visual inspection. The method includes learning, by a processor, group disentangled visual feature embedding vectors of input images. The input images include defective objects and defect-free objects. The method further includes generating, by the processor using a weight generation network, classification weights from visual features and semantic descriptions. Both the visual features and the semantic descriptions are for predicting defective and defect-free labels. The method also includes calculating, by the processor, a cosine similarity score between the classification weights and the group disentangled visual feature embedding vectors. The method additionally includes episodically training, by the processor, the weight generation network on the input images to update parameters of the weight generation network. The method further includes generating, by the processor using the trained weight generation network, a prediction of a test image as including any of defective objects and defect-free objects.

Deep group disentangled embedding and network weight generation for visual inspection

A false alarm reduction system and method are provided for reducing false alarms in an automatic defect detection system. The false alarm reduction system includes a defect detection system, generating a list of image boxes marking detected potential defects in an input image. The false alarm reduction system further includes a feature extractor, transforming each of the image boxes in the list into a respective set of numerical features. The false alarm reduction system also includes a classifier, computing as a classification outcome for the each of the image boxes whether the detected potential defect is a true defect or a false alarm responsive to the respective set of numerical features for each of the image boxes.

False alarm reduction system for automatic manufacturing quality control

A computer-implemented method and system are provided for training a model for New Class Categorization (NCC) of a test image. The method includes decoupling, by a hardware processor, a feature extraction part from a classifier part of a deep classification model by reparametrizing learnable weight variables of the classifier part as a combination of learnable variables of the feature extraction part and of a classification weight generator of the classifier part. The method further includes training, by the hardware processor, the deep classification model to obtain a trained deep classification model by (i) learning the feature extraction part as a multiclass classification task, and (ii) episodically training the classifier part by learning a classification weight generator which outputs classification weights given a training image.

Network reparameterization for new class categorization

A computer-implemented method executed by a processor for training a neural network to recognize driving scenes from sensor data received from vehicle radar is presented. The computer-implemented method includes extracting substructures from the sensor data received from the vehicle radar to define a graph having a plurality of nodes and a plurality of edges, constructing a neural network for each extracted substructure, combining the outputs of each of the constructed neural networks for each of the plurality of edges into a single vector describing a driving scene of a vehicle, and classifying the single vector into a set of one or more dangerous situations involving the vehicle.

Detecting dangerous driving situations by parsing a scene graph of radar detections

Growing community of inventors

Auburn Hills, MI, United States of America

Hans Peter Graf