We are presenting three papers in a row in tomorrow's Sight & Sound Workshop at CVPR'19. From event localization to explainable captioning and to cross-modal synthesizing, it's going to a feast for audio-visual modeling!!

Zooming Slow-Mo: Fast and Accurate One-Stage Space-Time Video Super-Resolution deepai.org/publication/zo… by Xiaoyu Xiang et al. including Yapeng Tian, Yulun Zhang, @chenliangxu #Interpolation #ComputerVision

Our Temporally-Deformable Alignment Network (TDAN) at #CVPR2020 restores a photo-realistic high-resolution video from its low-resolution version. Check out our poster at: rb.gy/unozi9.

Our one-stage Zooming Slow-Mo method at #CVPR2020 simultaneously increases temporal resolution and spatial resolution for an input video. Check out our poster here: rb.gy/thrav7

Our Deep Grouping Model (DGM) at #CVPR2020 combines a top-down segmentation process and a bottom-up grouping process for a unified perceptual scene parsing. The result is higher accuracy, better interpretability, and lower computation! Check out it here: rb.gy/xt7wsk

Our method learns a weakly-supervised video actor-action segmentation model with a wise selection of pseudo-annotations in iterative training. Its #CVPR2020 oral presentation is here: rb.gy/0mlg1a

URCS is highly ranked in Computer Vision for the Year 2020!

4 papers accepted by #CVPR2021. Congrats to my students and collaborators! Topics include audio-visual grounding-separation, audio-visual robustness, 3D video avatars, and language-driven image editing. Plus, we will co-organize a tutorial on Audio-Visual Scene Understanding.

One of our #CVPR2021 submissions got the final reviews: WA, WA, B, and it was rejected by the AC...

Welcome to join us this Saturday for our #cvpr2021 audio-visual scene understanding tutorial to learn recent advances in audio-visual learning. Website: …-visual-scene-understanding.github.io

Want to make computers that can see *and* hear? Come to the CVPR Sight and Sound workshop today! Schedule: sightsound.org Invited talks by: Justin Salamon, Chenliang Xu, Kristen Grauman, Dima Damen, Chuang Gan, John Hershey, Efthymios Tzinis, and James Traer

I have a postdoc opening (2 years) in continual and transfer learning from high-d imagery. Please contact me if you're interested. Start soon. #jobopening #ComputerVision #ai

I am a final-year PhD student actively looking for researcher and postdoc positions. My research focuses on trustworthy AI in computer vision, e.g., fairness and robustness. Let me know if you are interested. More information is on my homepage: zhiheng.li

Highly recommended my graduating PhD student, Zhiheng Li, who researches trustworthy issues in computer vision and has published numerous papers at top CV venues. If you have an opening for research scientist/post docs, please contact him!

Feel proud to hood three PhDs at a time. From left to right: Lele Chen Oppo US Research, Yapeng Tian UTD TT faculty, Jing Shi Adobe Research. Will meet again at Vancouver CVPR~

We will organize a virtual open house (on Nov 10 & 11). Learn all about our department, graduate programs, admissions, and more! Prospective PhD students are welcome to register it as soon as possible! Please Register here: t.ly/UyTp3 Likes and RT appreciated!

At #Neurips2023 next week to talk about our paper/AV-NeRF and all things AI, Audio/Speech & Multimodal😀 AV-NeRF:Learning Neural Fields for Real-World Audio-Visual Scene Synthesis Demo: t.ly/UuXkG Paper t.ly/0oIhV (w. Susan,Chao,Yapeng Tian,Chenliang Xu)

Can machines recording an audio-visual scene produce realistic, matching audio-visual experiences at novel positions and novel view directions? We present AV-NeRF to synthesize new videos with spatial audios along arbitrary novel camera trajectories in that scene. #NeurIPS2023

Is Open Source AI Bull? Open source AI is widely talked about. Yet, it doesn't have a clear definition. And, the status quo mostly falls short of any rigorous notion of openness. I wrote up an article with a principled definition of open source AI. medium.com/@jasoncorso/is…