Awesome

SVD Xtend

Stable Video Diffusion Training Code and Extensions 🚀

:bulb: Highlight

• Finetuning SVD. See Part 1.
• Tracklet-Conditioned Video Generation. Building upon SVD, you can control the movement of objects using tracklets(bbox). See Part 2.

Part 1: Training

Comparison

size=(512, 320), motion_bucket_id=127, fps=7, noise_aug_strength=0.00
generator=torch.manual_seed(111)

Init Image	Before Fine-tuning	After Fine-tuning

Video Data Processing

Note that BDD100K is a driving video/image dataset, but this is not a necessity for training. Any video can be used to initiate your training. Please refer to the DummyDataset data reading logic. In short, you only need to modify self.base_folder. Then arrange your videos in the following file structure:

self.base_folder
    ├── video_name1
    │   ├── video_frame1
    │   ├── video_frame2
    │   ...
    ├── video_name2
    │   ├── video_frame1
        ├── ...

Training Configuration(on the BDD100K dataset)

This training configuration is for reference only, I set all parameters of unet to be trainable during the training and adopted a learning rate of 1e-5.

accelerate launch train_svd.py \
    --pretrained_model_name_or_path=/path/to/weight \
    --per_gpu_batch_size=1 --gradient_accumulation_steps=1 \
    --max_train_steps=50000 \
    --width=512 \
    --height=320 \
    --checkpointing_steps=1000 --checkpoints_total_limit=1 \
    --learning_rate=1e-5 --lr_warmup_steps=0 \
    --seed=123 \
    --mixed_precision="fp16" \
    --validation_steps=200

Part 2: Tracklet2Video

Tracklet2Video

We have attempted to incorporate layout control on top of img2video, which makes the motion of objects more controllable, similar to what is demonstrated in the image below. The code and weights will be updated soon. It should be noted that we use a resolution of 512*320 for SVD to generate videos, so the quality of the generated videos appears to be poor (which is somewhat unfair to SVD), but our intention is to demonstrate the effectiveness of tracklet control, and we will resolve the issue with video quality as soon as possible.

Init Image	Gen Video by SVD	Gen Video by Ours

Methods

We have utilized the Self-Tracking training from Boximator and the Instance-Enhancer from TrackDiffusion. For more details, please refer to the paper.

:label: TODO List

• Support text2video (WIP)
• Support more conditional inputs, such as layout

:hearts: Acknowledgement

Our model is related to Diffusers and Stability AI. Thanks for their great work!

Thanks Boximator and GLIGEN for their awesome models.

:black_nib: Citation

If you find our work helpful for your research, please consider citing the following BibTeX entry.

@article{li2023trackdiffusion,
  title={Trackdiffusion: Multi-object tracking data generation via diffusion models},
  author={Li, Pengxiang and Liu, Zhili and Chen, Kai and Hong, Lanqing and Zhuge, Yunzhi and Yeung, Dit-Yan and Lu, Huchuan and Jia, Xu},
  journal={arXiv preprint arXiv:2312.00651},
  year={2023}
}