Open Source Linux AI Video Generators - Page 3

Sora.FM is positioned as a tool in the AI-generated video domain — likely aiming to let users produce video content via AI-driven workflows rather than classic manual editing. The project belongs to the growing class of “AI video generator / AI-assisted content creation” tools: it may use model-based generation, template-based editing, or combine video assets with generative models to automate parts of video creation or editing. For creators wanting to explore AI-based content generation — for example automated video clips, short-form media, or other generated video content — sorafm offers a starting point. As with many open-source generators in this space, the tradeoff lies in balancing ease-of-use and the limitations of generative output, but the fact that it’s publicly available means users can experiment, iterate, or fork to adapt pipelines: maybe customizing model prompts, video templates, or post-processing.

Implementation of Video Diffusion Models, Jonathan Ho's new paper extending DDPMs to Video Generation - in Pytorch. Implementation of Video Diffusion Models, Jonathan Ho's new paper extending DDPMs to Video Generation - in Pytorch. It uses a special space-time factored U-net, extending generation from 2D images to 3D videos. 14k for difficult moving mnist (converging much faster and better than NUWA) - wip. Any new developments for text-to-video synthesis will be centralized at Imagen-pytorch. For conditioning on text, they derived text embeddings by first passing the tokenized text through BERT-large. You can also directly pass in the descriptions of the video as strings, if you plan on using BERT-base for text conditioning. This repository also contains a handy Trainer class for training on a folder of gifs. Each gif must be of the correct dimensions image_size and num_frames.

Vidi is a family of large multimodal models developed for deep video understanding and editing tasks, integrating vision, audio, and language to allow sophisticated querying and manipulation of video content. It’s designed to process long-form, real-world videos and answer complex queries such as “when in this clip does X happen?” or “where in the frame is object Y during that moment?” — offering temporal retrieval, spatio-temporal grounding (i.e. locating objects over time + space), and even video question answering. Vidi targets applications like intelligent video editing, automated video search, content analysis, and editing assistance, enabling users to efficiently locate relevant segments and objects in hours-long footage. The system is built with open-source release in mind, giving developers access to model code, inference scripts, and evaluation pipelines so they can reproduce research results or integrate Vidi into their own video-processing workflows.

Wan Move is an open-source research codebase for motion-controllable video generation that focuses on enabling fine-grained control of motion within generative video models. It is designed to guide the temporal evolution of visual content by leveraging latent trajectory guidance, allowing users to manipulate how objects move over time without modifying the underlying generative architecture. By representing motion information as dense point trajectories and integrating them into the latent space of an image-to-video model, the project produces videos with more precise and controllable motion behavior than many existing methods. Wan-Move is particularly notable for eliminating the need for additional motion encoders, instead directly infusing motion cues into spatiotemporal features, which simplifies both training and inference.

video2robot is an end-to-end open-source pipeline that converts generative video or prompt-driven motion content into executable humanoid robot motion sequences, enabling researchers and developers to go from high-level action descriptions or videos to robot-ready motion data. The pipeline supports both prompt-to-video generation using models like Veo/Sora and video upload processing, followed by human pose extraction through a 3D pose model and retargeting of that motion to robot joints using a general motion retargeting system. This workflow allows users to generate robot motion files that specify joint angles, root positions, and orientations that can be deployed on supported robot platforms (e.g., Unitree models). Video2robot includes scripts for each stage of the pipeline (generation, extraction, conversion, visualization) and can run as a CLI or through a basic web UI.