Hugging Face
Models
Datasets
Spaces
Community
Docs
Enterprise
Pricing
Log In
Sign Up
1
7
5
Khalil Guetari
KhalilGuetari
Follow
YannisTevissen's profile picture
1 follower
Β·
8 following
AI & ML interests
None yet
Recent Activity
liked
a dataset
10 days ago
rghermi/sf20k
reacted
to
andito
's
post
with π₯
about 2 months ago
Many VLMs claim to process hours of video. But can they follow the story?π€ Today, we introduce TimeScope: The benchmark that separates true temporal understanding from marketing hype. Let's see how much VLMs really understand!β³ We test three skills that matter for real-world use: π Localized Retrieval: Find a specific action. π§© Information Synthesis: Piece together scattered clues. π Fine-Grained Perception: Analyze detailed motion (e.g., count how many times a person swings an axe). The results are in, and they're revealing. Only Gemini 2.5 pro handles 1-hour-long videos. Performance drops sharply with duration, proving that long video understanding is still challenging. We've found the breaking pointsβnow the community can start fixing them.π Want to learn more? TimeScope is 100% open-source. Benchmark your model and help us build the next generation of video AI. π Blog: https://huggingface.co/blog/timescope-video-lmm-benchmark π©βπ» Leaderboard & Demo: https://huggingface.co/spaces/Apollo-LMMs/TimeScope π Dataset: https://huggingface.co/datasets/Apollo-LMMs/TimeScope βοΈ Eval Code: https://github.com/EvolvingLMMs-Lab/lmms-eval
liked
a model
2 months ago
google/siglip2-base-patch16-512
View all activity
Organizations
Papers
3
arxiv:
2406.14938
arxiv:
2406.17590
arxiv:
2405.02305
spaces
1
Sleeping
First Agent Template
β‘
Find current time in any timezone
models
0
None public yet
datasets
0
None public yet