Video Generation Algorithm Researcher - Doubao Video Generation Large Model PixelDance

Team Introduction: We are the Doubao Video Generation Large Model-PixelDance team. The PixelDance model has industry-leading effects, precise semantic understanding capabilities, powerful dynamic effects and cool camera movements, and the industry's first consistent multi-lens generation, first and last frame control, which can be compatible with multiple styles and multiple proportions. Our team is committed to solving key problems in the field of video generation, with the absolute and ultimate SOTA goal, leading the future trend of technology. 1. Build a large-scale basic model for video generation and achieve a leading level 2. Design algorithms related to video generation and iterate the effects to improve the effect of video generation.

1. Experience in algorithm development in one or more fields, including video generation, image generation, multimodal pre-training, video understanding, and image understanding 2. Those who are familiar with the full process of video generation algorithm implementation are preferred, and those who have proposed Sota methods in related fields are preferred 3. Excellent coding skills, proficient in Pytorch, able to reproduce algorithms, and can optimize in a targeted manner 4. Excellent teamwork and continuous learning ability, and can efficiently promote project progress in teamwork.