Sora Model

At the 2024 Global Developer Pioneer Conference, the Moda community launched the "ModelScope-Sora Open Source Plan" to promote the exploration and innovation of China's Sora-like models through open source. The plan provides a one-stop tool chain, including data processing tools, multimodal data sets, Sora-like basic models, training and reasoning tools, etc. Moda has released the multimodal data processing system Data-Juicer, which contains more than 100 efficient operators, which can greatly improve the efficiency and quality of video data processing. Data-Juicer supports text, pictures,...

Tomcat announced today on its interactive platform that it is actively promoting the application for access to OpenAI's Sora model and plans to test the production of related materials. Currently, Tomcat has not officially accessed the Sora model. It is reported that Tomcat has been exploring the application of technology in the field of generated videos. The company has used tools such as Pika, Runway, and Stable Diffusion to produce some video materials and tried to apply them to marketing and animation production. However, Tomcat's application of Vincent video technology is still in the exploratory stage,…

According to media reports, Sora, the first video model released by OpenAI, has been very popular recently. Users can generate a one-minute high-definition video with just one sentence. Tesla CEO Elon Musk posted on the social media platform X, touting his company's simulated reality and video generation technology as the best. Musk said that Tesla mastered similar video generation technology a year ago, and the videos generated by these systems are not very interesting because the training data comes from Tesla cars. It looks like an ordinary video of Tesla, but in fact it is a dynamically generated world. Musk also pointed out that...