Inquiry about release plan for dataset, training code, and evaluation scripts

[tianciwu98-tech]

Hi authors:

Thank you for sharing this impressive work. I am very interested in reproducing the results and further studying the proposed omni-motion controlled multi-subject video customization framework.

I would like to ask whether there is a plan to release the following resources:

1. The curated training dataset
2. The DreamOmni Bench dataset / evaluation benchmark
3. Training code for Stage 1 omni-motion and identity supervised fine-tuning
4. Training code for Stage 2 latent identity reward feedback learning
5. Evaluation scripts for the metrics reported in the paper

Thanks again for your great work. Looking forward to any updates!