Why Is Video Still So Expensive? A Survey of Inference-Efficiency Mechanisms in Video and Audiovisual LLMs
Paper • 2609.10355 • Published • 18
Khalil Guetari
KhalilGuetari
AI & ML interests
None yet
Recent Activity
authored a paper 4 months ago
PEEK: Picking Essential frames via Efficient Knowledge distillation liked a Space 4 months ago
momentslab/peekOrganizations
upvoted a paper 13 days ago
authored a
paper 4 months ago
liked a
Space 4 months ago
upvoted a paper 4 months ago
updated 2
Spaces 5 months ago
liked a
Space 10 months ago
published a
Space 11 months ago
liked a
dataset about 1 year ago
reacted to andito's post with 🔥 about 1 year ago
Post
3142
Many VLMs claim to process hours of video. But can they follow the story?🤔
Today, we introduce TimeScope: The benchmark that separates true temporal understanding from marketing hype. Let's see how much VLMs really understand!⏳
We test three skills that matter for real-world use:
🔎 Localized Retrieval: Find a specific action.
🧩 Information Synthesis: Piece together scattered clues.
🏃 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 points—now the community can start fixing them.📈
Want to learn more? TimeScope is 100% open-source. Benchmark your model and help us build the next generation of video AI.
📖 Blog:
https://huggingface.co/blog/timescope-video-lmm-benchmark
👩💻 Leaderboard & Demo: Apollo-LMMs/TimeScope
📊 Dataset: Apollo-LMMs/TimeScope
⚙️ Eval Code: https://github.com/EvolvingLMMs-Lab/lmms-eval
Today, we introduce TimeScope: The benchmark that separates true temporal understanding from marketing hype. Let's see how much VLMs really understand!⏳
We test three skills that matter for real-world use:
🔎 Localized Retrieval: Find a specific action.
🧩 Information Synthesis: Piece together scattered clues.
🏃 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 points—now the community can start fixing them.📈
Want to learn more? TimeScope is 100% open-source. Benchmark your model and help us build the next generation of video AI.
📖 Blog:
https://huggingface.co/blog/timescope-video-lmm-benchmark
👩💻 Leaderboard & Demo: Apollo-LMMs/TimeScope
📊 Dataset: Apollo-LMMs/TimeScope
⚙️ Eval Code: https://github.com/EvolvingLMMs-Lab/lmms-eval
liked a
model about 1 year ago
reacted to sharpenb's post with 🔥 over 1 year ago
Post
3126
We open-sourced the
- Github repo: https://github.com/PrunaAI/pruna
- Documentation: https://docs.pruna.ai/en/stable/index.html
With open-sourcing, people can now inspect and contribute to the open code. Beyond the code, we provide detailed readme, tutorials, benchmarks, and documentation to make transparent compression, evaluation, and saving/loading/serving of AI models.
Happy to share it with you and always interested in collecting your feedback :)
pruna package that can be easily installed with pip install pruna :) It allows to easily ccompress and evaluate AI models including transformers and diffusers.- Github repo: https://github.com/PrunaAI/pruna
- Documentation: https://docs.pruna.ai/en/stable/index.html
With open-sourcing, people can now inspect and contribute to the open code. Beyond the code, we provide detailed readme, tutorials, benchmarks, and documentation to make transparent compression, evaluation, and saving/loading/serving of AI models.
Happy to share it with you and always interested in collecting your feedback :)
- 2 replies
reacted to prithivMLmods's post with 🔥 over 1 year ago
Post
2682
Gemma-3-4B : Image and Video Inference 🖼️🎥
🥠Git : https://github.com/PRITHIVSAKTHIUR/Gemma-3-Multimodal
@gemma3 : {Tag + Space_+ 'prompt'}
@video-infer : {Tag + Space_+ 'prompt'}
+ Gemma3-4B : google/gemma-3-4b-it
+ By default, it runs : prithivMLmods/Qwen2-VL-OCR-2B-Instruct
Gemma 3 Technical Report : https://storage.googleapis.com/deepmind-media/gemma/Gemma3Report.pdf
🥠Git : https://github.com/PRITHIVSAKTHIUR/Gemma-3-Multimodal
@gemma3 : {Tag + Space_+ 'prompt'}
@video-infer : {Tag + Space_+ 'prompt'}
+ Gemma3-4B : google/gemma-3-4b-it
+ By default, it runs : prithivMLmods/Qwen2-VL-OCR-2B-Instruct
Gemma 3 Technical Report : https://storage.googleapis.com/deepmind-media/gemma/Gemma3Report.pdf
- 1 reply
liked a
Space over 1 year ago
liked a
dataset over 1 year ago
updated a
Space over 1 year ago
reacted to cfahlgren1's post with 👀 almost 2 years ago
Post
1165
If you are like me, I like to find up and coming datasets and spaces before everyone else.
I made a trending repo space cfahlgren1/trending-repos where it shows:
- New up and coming Spaces in the last day
- New up and coming Datasets in the last 2 weeks
It's a really good way to find some new gems before they become popular. For example, someone is working on a way to dynamically create assets inside a video game here: gptcall/AI-Game-Creator
I made a trending repo space cfahlgren1/trending-repos where it shows:
- New up and coming Spaces in the last day
- New up and coming Datasets in the last 2 weeks
It's a really good way to find some new gems before they become popular. For example, someone is working on a way to dynamically create assets inside a video game here: gptcall/AI-Game-Creator
reacted to fdaudens's post with 👍 almost 2 years ago
Post
2371
🔍 NYT leveraged AI to investigate election interference by analyzing 400+ hours of recorded meetings - that's 5M words of data!
AI spotted patterns, humans verified facts. Every AI-flagged quote was manually verified against source recordings. Really appreciate that they published their full methodology - transparency matters when using AI in journalism.
A perfect blend of tech & journalism.
The future of journalism isn't robots replacing reporters - it's AI helping humans process massive datasets more efficiently. Sometimes the most powerful tech solutions are the least flashy ones.
Read the article: https://www.nytimes.com/interactive/2024/10/28/us/politics/inside-the-movement-behind-trumps-election-lies.html?unlocked_article_code=1.Vk4.ucv9.dbHVquTQaf0G&smid=nytcore-ios-share
AI spotted patterns, humans verified facts. Every AI-flagged quote was manually verified against source recordings. Really appreciate that they published their full methodology - transparency matters when using AI in journalism.
A perfect blend of tech & journalism.
The future of journalism isn't robots replacing reporters - it's AI helping humans process massive datasets more efficiently. Sometimes the most powerful tech solutions are the least flashy ones.
Read the article: https://www.nytimes.com/interactive/2024/10/28/us/politics/inside-the-movement-behind-trumps-election-lies.html?unlocked_article_code=1.Vk4.ucv9.dbHVquTQaf0G&smid=nytcore-ios-share
upvoted an article about 2 years ago
Article
FineVideo: behind the scenes
- +4
mfarre, andito, lewtun, lvwerra, pcuenq, thomwolf
• • 35 authored a
paper about 2 years ago