This AI article presents InstructVideo: a new AI approach to improve text-to-video dissemination models using human feedback and efficient tuning techniques

Broadcast models have become the predominant approach to generating videos. However, its reliance on large-scale web data, which varies in quality, often leads to results that lack visual appeal and do not align well with the textual cues provided. Despite recent advances, there is still room to improve the visual quality of the generated videos. A notable factor contributing to this challenge is the varying quality of the extensive web data used in pre-training. This variability can result in models capable of producing content that lacks visual appeal, may be toxic, and does not align well with the provided prompts.

A team of researchers from Zhejiang University, Alibaba Group, Tsinghua University, Singapore University of technology and Design, S-Lab, Nanyang Technological University, CAML Laboratory and Cambridge University presented InstructVideo to instruct models text-to-video broadcasting with humans. feedback by adjusting rewards. Comprehensive experiments, encompassing qualitative and quantitative evaluations, confirm the practicality and effectiveness of incorporating image reward models into InstructVideo. This approach significantly improves the visual quality of the generated videos without compromising the model's ability to generalize.

Early efforts in video generation focused on GANs and VAEs, but generating videos from text remained a challenge. Broadcast models have emerged as the de facto method for video generation, providing diversity and fidelity. VDM extended image diffusion models to video generation. Efforts were made to introduce spatiotemporal conditions for a more controllable generation. Understanding human preferences in visual content generation is challenging, and some work uses annotations and fitting of annotated data to model human preferences. Learning from human feedback is desirable in optical content generation, and previous work focused on reinforcement learning and agent alignment.

InstructVideo uses a reformulation of reward tuning as editing, improving computational efficiency and effectiveness. The method incorporates segmental video reward (SegVR) and temporally attenuated reward (TAR) to enable efficient reward tuning using image reward models. SegVR provides reward signals based on segmental sparse sampling, while TAR mitigates temporal modeling degradation during fine-tuning. The optimization objective is rewritten to include the degree of the attenuation rate, with a default value of 1 for the coefficient. InstructVideo takes advantage of the diffusion process to obtain the starting point for adjusting rewards in video generation.

The research provides more visualization results to exemplify the conclusions drawn, showing how the generated videos evolve. The effectiveness of InstructVideo is demonstrated by an ablation study in SegVR and TAR, showing that its removal leads to a noticeable reduction in temporal modeling capabilities. InstructVideo consistently outperforms other methods in terms of video quality, with improvements in video quality being more pronounced than improvements in text and video alignment.

In conclusion, the InstructVideo method significantly improves the visual quality of the generated videos without compromising generalization capabilities, as validated through extensive qualitative and quantitative experiments. InstructVideo outperforms other methods in terms of video quality, with improvements in video quality being more pronounced than improvements in video-text. alignment. Using image reward models, such as HPSv2, in InstructVideo is practical and effective in improving the visual quality of the generated videos. Incorporating SegVR and TAR into InstructVideo improves tuning and mitigates temporal modeling degradation.

Review the Paper. All credit for this research goes to the researchers of this project. Also, don't forget to join. our SubReddit of more than 35,000 ml, 41k+ Facebook community, Discord channel, and Electronic newsletterwhere we share the latest news on ai research, interesting ai projects and more.

If you like our work, you'll love our newsletter.

<!– ai CONTENT END 2 –>