Valentin Gabeur (original) (raw)

Valentin Gabeur AI researcher at Google DeepMind (Gemini), specializing in multi-modal learning. Postdoc at Meta AI (SAM 2), PhD from Inria, MSc from Toulouse University, MSc from Icam. Former mechanical engineer with 6 years of experience in industrial automation. Email / CV / Google Scholar / LinkedIn / Twitter / GitHub

	SAM 2: Segment Anything in Images and Videos Nikhila Ravi, Valentin Gabeur, Yuan-Ting Hu, Ronghang Hu, Chaitanya Ryali, Tengyu Ma, Haitham Khedr, Roman Rädle, Chloe Rolland, Laura Gustafson, Eric Mintun, Junting Pan, Kalyan, Vasudev Alwala, Nicolas Carion, Chao-Yuan Wu, Ross Girshick, Piotr Dollár, Christoph Feichtenhofer* arXiv, 2024 paper / demo / project page / blog / dataset / github Promptable visual segmentation in images and videos.
	AVATAR: Unconstrained Audiovisual Speech Recognition Valentin Gabeur, Paul Hongsuck Seo, Arsha Nagrani, Chen Sun, Karteek Alahari, Cordelia Schmid INTERSPEECH, 2022* arXiv / project page / bibtex Leveraging the full frame visual context to improve speech recognition in videos.
	Masking Modalities for Cross-modal Video Retrieval Valentin Gabeur, Arsha Nagrani, Chen Sun, Karteek Alahari, Cordelia Schmid WACV, 2022 arXiv / bibtex Pre-training strategy for learning multi-modal fusion from unlabelled videos.
	Multi-modal Transformer for Video Retrieval Valentin Gabeur, Chen Sun, Karteek Alahari, Cordelia Schmid ECCV, 2020 (Spotlight paper) arXiv / code, models, data / bibtex Cross-modal architecture to encode language captions and videos in a common embedding space.
	CVPR 2020 Video Pentathlon Challenge: Multi-modal Transformer for Video Retrieval Valentin Gabeur, Chen Sun, Karteek Alahari, Cordelia Schmid CVPR Video Pentathlon Workshop, 2020 (First place) report /paper /challenge /recording Winning approach for the CVPR 2020 Video Pentathlon Challenge, a video retrieval competition.
	Moulding Humans: Non-parametric 3D Human Shape Estimation from Single Images Valentin Gabeur, Jean-Sebastien Franco, Xavier Martin, Cordelia Schmid, Gregory Rogez ICCV, 2019 arXiv / bibtex Efficient 3D shape representation through the combination of depth maps.