International Business Machines Corporation

Massachusetts Institute of Technology

Indian Institute of Technology

Rameswar Panda

Rogerio Schmidt Feris

Richard Chen

Quanfu Fan

Chuang Gan

Bo Wu

Michele Merler

Aude Jeanne Oliva

Wei Zhang

Bishwaranjan Bhattacharjee

Ulrich Alfons Finkler

Dakuo Wang

Pin-Yu Chen

Sijia Liu

Avirup Sil

Hui Wu

Taesun Moon

Chung-Ching Lin

Mayoore Selvarasa Jaiswal

Yue Meng

David Cox

Aashka Trivedi

Bowen Pan

Abir Das

Yi Li

Yoon Hyung Kim

Methods, systems, and computer program products for translating text using generated visual representations and artificial intelligence are provided herein. A computer-implemented method includes generating a tokenized form of at least a portion of input text in a first language; generating at least one visual representation of at least a portion of the input text using a first set of artificial intelligence techniques; generating a tokenized form of at least a portion of the at least one visual representation; and generating an output including a translated version of the input text into at least a second language by processing, using a second set of artificial intelligence techniques, at least a portion of the tokenized form of the at least a portion of the input text and at least a portion of the tokenized form of the at least a portion of the at least one visual representation.

Translating text using generated visual representations and artificial intelligence

An embodiment includes identifying an initial plurality of sets of hyperparameter values at which to evaluate an objective function that relates hyperparameter values to performance values of a neural network. The embodiment also executes training processes on the neural network with the hyperparameters set to the each of the initial sets of hyperparameter values such that the training process provides an initial set of the performance values for the objective function. The embodiment also generates an approximation of the objective function using splines at selected performance values. The embodiment approximates a point at which the approximation of the objective function reaches a maximum value, then determines an updated set of hyperparameter values associated with the maximum value. The embodiment then executes a runtime process using the neural network with the hyperparameters set to the updated set of hyperparameter values.

Configuring a neural network using smoothing splines

A neural architecture search method, system, and computer program product that determines, by a computing device, a best fit language model of a plurality of language models that is a best fit for interpretation of a corpus of natural language and interprets, by the computing device, the corpus of natural language using the best fit language model.

Neural architecture search of language models using knowledge distillation

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 sequence of patch tokens representing an image can be received. A network can be trained to learn informative patch tokens and uninformative patch tokens in the sequence of patch tokens, in learning to recognize an object in the image. The sequence of patch tokens can be reduced by removing the uninformative patch tokens from the sequence of patch tokens. The reduced sequence of patch tokens can be input to an attention-based deep learning neural network. The attention-based deep learning neural network can be fine-tuned to recognize the object in the image using the reduced sequence of patch tokens.

Interpretability-aware redundancy reduction for vision transformers

A base pathway of a computerized two-pathway video action recognition model is trained using a plurality of labeled video samples. The base pathway is trained using a plurality of unlabeled video samples at a first framerate. An auxiliary pathway of the computerized two-pathway video action recognition model is trained using a plurality of the unlabeled video samples at a second framerate, the second framerate being slower than the first framerate, wherein the training of the base pathway and the training of the auxiliary pathway result in a trained computerized two-pathway video action recognition model. A candidate video is categorized using the trained computerized two-pathway video action recognition model and the categorized candidate video is stored in a computer-accessible video database system for information retrieval.

Temporal contrastive learning for semi-supervised video action recognition

Methods, apparatus, and systems for multi-resolution processing for video classification. A plurality of video frames of a video are obtained and a resolution for classifying each video frame of the plurality of video frames is determined by analyzing each video frame using a policy network. Based on the determined resolution, each video frame having a determined resolution is rescaled and each rescaled video frame is routed to a classifier of a backbone network that corresponds to the determined resolution. Each rescaled video frame is classified using the corresponding classifier of the backbone network to obtain a plurality of classifications and the classifications are averaged to determine an action classification of the video.

Dynamic multi-resolution processing for video classification

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

Machine learning-based techniques for summarizing collections of data such as image and video data leveraging side information obtained from related (e.g., video) data are provided. In one aspect, a method for video summarization includes: obtaining related videos having content related to a target video; and creating a summary of the target video using information provided by the target video and side information provided by the related videos to select portions of the target video to include in the summary. The side information can include video data, still image data, text, comments, natural language descriptions, and combinations thereof.

Summarizing videos via side information

A sequence of frames of a video can be received. For a given frame in the sequence of frames, a gradient-embedded frame is generated corresponding to the given frame. The gradient-embedded frame incorporates motion information. The motion information can be represented as disturbance in the gradient-embedded frame. A plurality of such gradient-embedded frames can be generated corresponding to a plurality of the sequence of frames. Based on the plurality of gradient-embedded frames, a neural network such as a generative adversarial network is trained to learn to suppress the disturbance in the gradient-embedded frame and to generate a substitute frame. In inference stage, anomaly in a target video frame can be detected by comparing it to a corresponding substitute frame generated by the neural network.

Growing community of inventors

Medford, MA, United States of America

Rameswar Panda