Julien Despois

ML Researcher. Strongly interested in rationality, epistemology, alignment, as well as generative models, their shortcomings and societal implications.

[Robin KIPS @robin_kips]

Our Generative Legs technology is now available for all Quest devices !

[Julien Despois @DespoisJ]

@OHamzaoui1 @jadengis I wonder in which book you read that! ;)

[Julien Despois @DespoisJ]

@sigma__dev If you do, the other bars will be longer too, and the improvement between the different methods will still be the same. Plus the figures are provided. But the better way would be to visually indicate with dashes that the left part of the bars is truncated.

[Mario Klingemann💧💦 @quasimondo]

Sounds like it is just a matter of time until someone ports Doom for fruit-fly OS.

[AK @_akhaliq]

Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking Inputs with Diffusion Model

abs: https://t.co/7PukJpWw2V
project page: https://t.co/q5NzKS9us5

[Yuvi @yvrjsharma]

🎉You have seen ControlNet's magic, now witness the power of grounded image generation using the state-of-the-art 💥GLIGEN (CVPR2023)💥

🔥This incredible Gligen webui is powered by @Gradio, hosted on 🤗Spaces, and is going brrr with community grant GPUs.🤑

😍Demo video - I generated an image of 'John Lennon using a PC' and then had fun augmenting it by creating subsequent bounding boxes and image generation.

Acess the mind-blowing @huggingface Space here - https://t.co/wcHb9gRqpu

[Amina ✨ @AminaZakhnouf_]

Aidez nous à lever assez pour relever le débat ! :p (pardon je ne recommencerais plus). Plus sérieusement on porte ce projet à bout de bras depuis un an, aidez nous à passer à l’échelle et faire de la politique publique l’affaire de TOUS !

[Robin KIPS @robin_kips]

Glad to annouce that our latest paper "Deep Graphics Encoder for Real-Time Video Makeup Synthesis from Example" will be presented at CVPR Workshop AICC and CVFAD.
https://t.co/7syQvdEXXt

#CVPR #ML #AR

[arXiv Daily @Arxiv_Daily]

Learning Long-Term Style-Preserving Blind Video Temporal Consistency
https://t.co/Bi0p13R8aS
by Hugo Thimonier et al. including @robin_kips, @MatthieuPerrot
#RecurrentNeuralNetwork #ComputerVision

[Julien Despois @DespoisJ]

@quasimondo If you want to be petty, post the encrypted idea on Twitter for posterity, and tweet the decryption key when you get scooped 😅

[Julien Despois @DespoisJ]

@karpathy On my default install, it happens when I type "x" but if I type "z" (not ctrl + z), the cell comes back.

[Julien Despois @DespoisJ]

My first Deep Learning paper on HD Face Aging has been accepted at the AIM workshop of ECCV 2020!

Presentation video: https://t.co/GyNGlxagPE
Project page: https://t.co/MxlPAkcQUO
pdf: https://t.co/nKvtHYIcSJ
abs: https://t.co/DqMWMdziBG

#ai #deeplearning #machinelearning #gan

[Robin KIPS @robin_kips]

Checkout our paper on makeup synthesis using GANs, curently presented at the Advanced in Image Manipulation workshop of ECCV.

Project page: https://t.co/1d1ZyIOwvh
ArXiv: https://t.co/FuiDHfoL2v
Presentation video : https://t.co/PK9ut5tXps

#AI #GAN #ECCV20 #AIM20 #ML

[AK @_akhaliq]

AgingMapGAN (AMGAN): High-Resolution Controllable Face Aging
project page: https://t.co/2QdTWzYtvW
video: https://t.co/NgTTaNM4Hp

[AK @_akhaliq]

CA-GAN: Weakly Supervised Color Aware GAN for Controllable Makeup Transfer
pdf: https://t.co/koLLPww9nM
abs: https://t.co/f7EJiC44iU
project page: https://t.co/xWiA4QVV1t

[Julien Despois @DespoisJ]

@elonmusk Shouldn’t the reflection be warped by the spherical shape of the glass?

[Julien Despois @DespoisJ]

@InertialObservr @tenortrombone No I think he is talking about the other meta-stable positions

[Julien Despois @DespoisJ]

@quasimondo I’m curious how you define ‘real time’ for movement in a latent space. 🤔

[Fritz AI @fritzlabs]

In part 2 of his excellent series looking at generative #MachineLearning models on-device, @DespoisJ looks at what they might be able to do for text- and audio-based applications. https://t.co/bOMDeGxvnS