🎯 Modeling how people interact with objects in #3D is important for both 3D #perception from 2D images, and for 3D #avatar synthesis for AR/VR 👉 I gave a talk on this @ #BMVA 'Symposium on Digital Humans' 👉 Video: https://t.co/RJs3fOygDX 👉 Slides: https://t.co/8wJwEGYuss

📢 Working on Digital Humans/Avatars? Register for the BMVA Symposium on Digital Humans on the 28th of May in London! We're also looking for recent work to be presented on the day. Not sure if your work fits? Send me a DM. Register/submit here 👉 https://t.co/nXHfAevacd Many

I'm excited to share what I've been working on during my summer internship at @Microsoft . GASP creates photorealistic, real-time Avatars from an image or short video. Project page: https://t.co/oW3oh4UxH5 Arxiv paper: https://t.co/Y60X5mn66Z Demo Video: https://t.co/E4y187MmYM

Anne Gagneux, Ségolène Martin, @qu3ntinb, Remi Emonet and I wrote a tutorial blog post on flow matching: https://t.co/TSkg1VZ5cn with lots of illustrations and intuition! We got this idea after their cool work on improving Plug and Play with FM: https://t.co/0GRNZd3l8O

The Computer Vision Foundation has just posted all of the recorded talks from #ICCV2023 up on its YouTube channel. Check them out here: https://t.co/hoOc0vqQKf

🚀 Exciting Research Alert! Traditional #AIAlignment #RLHF methods are expensive & require updating billions of parameters. 🔥 Is it possible to do #LLMAlignment without finetuning model parameters? ✅ YES! Transfer Q*: Principled Decoding Alignment https://t.co/oCiOxIy5VS

This might be one of the most important 45-mn read you could indulge in today if you want to understand the secret behind high performance large language models like Llama3, GPT-4 or Mixtral Inspired by the @distillpub interactive graphics papers, we settled to write the most

An important update on the #CVPR2024 submission deadline from the conference organizing committee: Our Program Chairs have voted to shift the CVPR 2024 submission deadline to November 17th (A one-week extension). The website will be updated shortly to reflect this change.

I just published How to read a research paper: 5 tips to reading effectively. Hope it is useful for some persons.

Super excited to share that my first-ever paper, READ Avatars, has been accepted to @BMVCconf. Project Page: https://t.co/8WPCSLFeoe Blog post: https://t.co/ZaC6SxzKHd TLDR: We bring emotion into audio-driven talking head generation using neural rendering.

This is what happens when you don't chase trends and jump at shiny things.

Last week my university @UGent approved non-competitive research financing for all professors. The idea is that each professor (all ranks) would receive a yearly amount of research funding of ca. 30.000 euro (so two professors can jointly hire one PhD student). 1/2

🚨Here’s an intuitive explanation for why training on lots and lots of data creates emergent properties, for instance math and reasoning, in large language models like #GPT-4 and #ChatGPT 🚨 1/17

How to do experiments? Junior students often feel stressed before the weekly meeting with their advisors because their experiments do not go well. 😩😰😱 Some tips on why, what, and how to do experiments. 🧵

Neural network trainings are nondeterministic. Repeated runs each produce a unique network, often with significantly _varying_ test-set performance. 🆕📜 I demonstrate that this variation has a simple statistical structure, and is harmless & inevitable https://t.co/1zzpNHi0Vy

A true success story in computing without a fancy name.

I asked #Galactica about some things I know about and I'm troubled. In all cases, it was wrong or biased but sounded right and authoritative. I think it's dangerous. Here are a few of my experiments and my analysis of my concerns. (1/9)

Joint work with @vinaypn and @skymanaditya1

INR-V: A Continuous Representation Space for Video-based Generative Tasks Bipasha Sen, Aditya Agarwal, Vinay P Namboodiri, C.V. Jawahar https://t.co/x6GyGzSraI

We have seen extensive work on "Image Inversion"; but what is "Video Inversion?". In our latest work, INR-V, accepted at @TmlrOrg, we propose a novel video representation space that can be used to invert videos (complete and incomplete!). Project page: https://t.co/u1jVj3e4z5