jiangzeyinzi

We have received feedback regarding the color shift issue. Currently, it appears to be a minor point that has not been thoroughly researched and discussed in the video editing field,...

@pftq Would trying different seeds in each editing operation help alleviate the issue? @hoveychen This case is quite intuitive. Could you compare whether there is a shift issue when encoding...

@pftq I am only providing a possible clue for alleviation, which is to use different seeds for each round when generating segments of long videos; this may help. Currently, the...

Currently, we only support Linux environments and have not yet validated on the Windows environment. We will conduct relevant verifications in the future. We also recommend using the limited-time free...

Please check this #8

Based on your description, I'm still a bit confused about which task you would like to execute: 1) Conditional generation task: Generate an image from a caption + segmentation map....

The scepter_img2img_hed_example YAML or data format you are using is supported by another work, stylebooth, for image editing tasks. In SCEdit, we extract image conditions through on-the-fly operation by registering...

The training strategies for the 1.3B and 14B models are the same, so there is no distinction in functionality based on scale. Please try different seeds or provide complete input...

The VACE model has only undergone `Context Adapter Tuning` and has not made any parameter changes to the original Wan2.1-T2V-1.3B/14B. This design converges more quickly without compromising the capabilities of...

We recommend that users modify [vace_wan_inference.py](https://github.com/ali-vilab/VACE/blob/main/vace/vace_wan_inference.py) themselves to support sequential batch inference for multiple data entries.