🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Few fav papers from sess8,9,10 (phew hard to catch up!) 1/ "Playable Video Generation" Sergey Tulyakov & team. Now you can control Djokovic playing tennis! Cool idea of estimating action states and then generating target image for that action state from a tennis video.

Congrats to Willi Menapace for receiving the Alfredo Petrosino Award as Best Master Thesis from the Italian Association for Computer Vision, Pattern Recognition and Machine Learning. It was a pleasure to work with you and Stéphane LATHUILIÈRE ! UniTrento_DISI #cvpl

⚡️Lightning Spotlight⚡️ solo-learn: An easy-to-use library and a reproducible benchmark for assessing #selfsupervised representation learning methods. 👏🏽 Kudos Enrico Fini & Victor Turrisi!

I am looking for a new PhD student to join my research group and work on Tiny Machine Learning. Send me a DM or email if you are interested! #research #phd #job_opportunities #jobposting

Honourable Mention ICCV2021 Viewing Graph Solvability via Cycle Consistency Federica Arrigoni (University of Trento), Andrea Fusiello, Elisa, Tomas Pajdla [Session 5 A/B]

"Quantum Multi-Model Fitting ", #CVPR2023 (Highlight). Our formulation can be efficiently sampled by a quantum annealer without the relaxation of the objective. We propose iterative and decomposed versions of QuMF. Draft: arxiv.org/pdf/2303.15444… Code: github.com/FarinaMatteo/q…

2. Want to generate the whole city in 3D? Checkout InfiniCity - a method that does exactly that! Project: hubert0527.github.io/infinicity/ with Chieh Hubert Lin (Job Hunting For 2025) Hsin-Ying Lee Willi Menapace Menglei Chai, Aliaksandr Siarohin, Ming-Hsuan Yang and yours truly.

Excited to share our latest work 'Snap Video'! A great collaboration with Aliaksandr Siarohin Ivan Skorokhodov Kate Deyneka Tsai-Shien Chen Anil Kag Yuwei Fang Aleksei Stoliar Elisa Ricci Jian Ren Sergey Tulyakov More to come soon! Project page: snap-research.github.io/snapvideo/

📢📢📢 𝐀𝐂𝟑𝐃: Analyzing and Improving 3D Camera Control in Video Diffusion Transformers snap-research.github.io/ac3d TL;DR: for 3D camera control in generative video, it really helps knowing *which* part of your model you should mess with Internship by Sherwin Bahmani at Snap

📢MinT: Temporally-Controlled Multi-Event Video Generation📢 mint-video.github.io TL;DR: We identify a fundamental failure mode of existing video generators: they cannot produce videos with sequential events. MinT unlocks this capability with temporal grounding of events. 🧵

MinT beats Sora in multi-event generation! One week after the release of MinT, Sora also released a *storyboard* tool that targets the same task (sequential events + time control). Below are a few comparisons, where MinT shows better event transition and timing: (1/N)

Video-to-Audio and Audio-to-Video models struggle with temporal alignment. AV-Link solves the problem by conditioning on diffusion model features Great collaboration with Moayed Haji Ali , Aliaksandr Siarohin , Ivan Skorokhodov , Alper Canberk , Kwot Sin Lee, Vicente Ordonez and Sergey Tulyakov

Check out Video Alchemist Our latest work enables Multi-subject open-set personalization with no need for inference-time tuning 👇👇👇

What if you could compose videos— merging multiple clips, even capturing complex athletic moves where video models struggle - all while preserving motion and context? And yes, you can still edit them with text after! Stay tuned for more results. #AI #VideoGeneration #SnapResearch

We discovered that imposing a spatio-temporal weight space via LoRAs on DIT-based video models unlocks powerful customization! It captures dynamic concepts with precision and even enables composition of multiple videos together!🎥✨

In the past 1.5 weeks, there appeared 2 papers by 2 different research groups which develop the exactly same (and embarrassingly simple) trick to improve convergence of image/video diffusion models by 20-100+% (sic!) arxiv.org/abs/2502.14831 arxiv.org/abs/2502.09509