Thu Dec 08 2022
Tue Dec 06 2022

Unifying Vision, Text, and Layout for Universal Document Processing

Natural Language Processing
Document AI
Document understanding and QA
Neural document editing
Content customization

Sets the SotA on 9 Document AI tasks, e.g., document understanding and QA, across diverse data domains. Ranks first on the Document Understanding Benchmark (DUE).

Can improve document understanding and QA across diverse data domains such as finance reports, academic papers, and websites. Also allows for high-quality neural document editing and content customization.

https://arxiv.org/pdf/2212.02623.pdf
https://arxiv.org/abs/2212.02623
https://twitter.com/arankomatsuzaki/status/1600308880374104064/photo/1

InternVideo: General Video Foundation Models via Generative and Discriminative Learning

Deep Learning
Computer Vision
Video understanding
Video action recognition/detection
Video-language alignment

Presents general video foundation models, InternVideo, by taking advantage of both generative and discriminative self-supervised video learning.

Can improve video understanding for various tasks including video action recognition/detection, video-language alignment, and open-world video applications.

https://arxiv.org/pdf/2212.03191.pdf
https://arxiv.org/abs/2212.03191
https://twitter.com/arankomatsuzaki/status/1600307950068715520/photo/1
Mon Dec 05 2022
Sun Dec 04 2022
Tue Nov 29 2022
Wed Nov 23 2022