Comments (2)
@junchen14 Thanks for your question and awesome work! The main difference is shown in the name of the two projects:
- Video ChatCaptioner and ChatCaptioner focus on generating better and fine-grained captions automatically. They use ChatGPT as a smart questioner, while using BLIP2 as a good answerer.
- Our project aims to build a chatbot to ask anything about video. It's more like your team's other project miniGPT-4. Different from miniGPT-4 and BLIP2, which implicitly encode the image into abstract tokens, we currently adopt explicit encoding (frame caption, dense caption with box, action class...). Without extra training, we find that the explicit encoding is better now. BTW, we also try to simply adapt miniGPT-4 for video. Unfortunately, without instruction fine-tuning, the results are sometimes unsatisfactory now.
from ask-anything.
Thanks for the clarification! It looks like an interesting project and glad to see that miniGPT-4 for video has also been working in progress!
from ask-anything.
Related Issues (20)
- get_text_encoder is None in VideoChat2 HOT 1
- video chat2 dataset cc3m path HOT 1
- 输出文本长度偏短 HOT 2
- urgent problem HOT 1
- how to treat video datasets differently with the same as paper? HOT 3
- RuntimeError: checkpoint url or path is invalid HOT 1
- no info about kwai-videoLLM on the MVBench Leaderboard HOT 1
- Instruction tuning with my own datasets HOT 4
- Fail to run v2 with flash attention HOT 4
- VideoChat2第三阶段的预训练模型使用 HOT 3
- vido chat2在线链接 HOT 1
- sequential ` for samples in batch` loop in Videochat2 is slow, any ways to parallalize it ? HOT 1
- 34b large language model hermes2_yi34b HOT 5
- VideoChat2: release any other video/image encoders? HOT 1
- VideoChat2, 仅finetune stage3后,视频描述存在大量重复 HOT 3
- flash_attn版本讨论 HOT 3
- VideoChat2运行疑问 HOT 3
- meta官方好像已经删除了llama1的权重,请问从哪里还可以获得llama1的huggingface版本呢? HOT 2
- 一二阶段训练数据读入问题 HOT 1
- Question about training stage3 in videochat2 HOT 3
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from ask-anything.