Request for adding new MLLM benchmark: OmniBench

[yizhilll]

Hi,

Nice work on indexing useful papers in this repo!

My team has released a new benchmark for the ability for "omni understanding" the image, audio and text of the MLLMs. We are quite confident that this could be a useful resource for the recent research in the field, and sincerely hope that we could share our work in this repo, thanks!

links:

• project page https://m-a-p.ai/OmniBench/
• paper: https://arxiv.org/abs/2409.15272

Best

[xjtupanda]

Thanks for your sharing! We've added your work to the repo.