Haiwen Diao

icon indicating a website link Paranioar.github.io/

icon company Nanyang Technological University icon location Research Topic: Vision-and-Language, Parameter/Memory-Efficient Transfer Learning, Large Multi-Modality Models for Perception and Generation.

Results 6 comments of Haiwen Diao

Trained model

> Is there the piece of code and trained model available in public domain open source? At present, only the unique code links of each paper are collected. In the...

Will this tutorial continue to be updated?

> _No description provided._ Yes. Due to some affairs, some papers have not been updated until the middle of April.

Will this tutorial continue to be updated?

> _No description provided._ Since energy is limited, I welcome readers to update this tutorial by providing papers I missed. Thanks very much！

Add new paper pls

> Hi Paranioar! Could you please include this recent work about video-language pretraining and noisy correspondence? Thanks a lot. > > **(ICLR2024_Norton) Multi-granularity Correspondence Learning from Long-term Noisy Videos** _Yijie...

Some questions about encoder-free VLMs

> How wonderful the Work Eve series is!!! I have the following questions that I would like to ask you, and I would be extremely grateful if you could provide...

Finetune on downstream tasks

Good questions. For object localization tasks, we suggest directly outputting the bounding box coordinates. Two important notes are: **(1) Replace the original 1D RoPE with 2D RoPE** to better capture...