Shulou(Shulou.com)12/24 Report--

On December 20, Liad Virtual moving Point officially released the "LYDIA" action model on AI results release Day. Huawei Cloud, JD.com Technology, iFLYTEK, Tianyi Cloud, Intel, Amazon Cloud and other partners attended the press conference to witness the historical moment of the AI era.

CMO of Liad Group and Liu Yaodong, CEO of Virtual Motion Point, released the LYDIA action model on the spot, which is a professional ability model independently developed by Liad Virtual Motion Point. It has the core abilities of "cognition, perception, prediction and generation" of spatial action data, and takes the lead in exploring the new model of AIGC in the field of global spatial computing and action generation.

Based on the accumulation of action data

Bow into the field of big model

The bottleneck of research and development of large models in the AI era is no longer the development of algorithms, but the accumulation of data. Liad Virtual Motion Point has accumulated a large amount of motion data in the field of spatial computing and motion capture for many years, which is unmatched by other AI large model companies, and it is also the core reason why Liad Virtual Motion Point can develop large motion models.

The parameters of the LYDIA action model include two parts: the basic data parameters and the core action data parameters. In terms of basic data parameters, LYDIA refers to today's mainstream large model data sets, and reaches the scale of tens of billions of sets in the process of training the model for natural language understanding; in terms of core action data parameters, it has many years of data accumulation of optitrack optical dynamic capture, data precipitation of unmarked point dynamic capture, and action data extraction in Internet public videos. With the accumulation of three kinds of action data, Liad virtual moving point successfully carried out cognitive training on the action data of the large LYDIA model.

LYDIA action model

Don't write poetry, just do things, and be more focused.

Liu Yaodong said in the sharing: "LYDIA is one of the few large models around the world that really achieve action generation." Different from other general large models, LYDIA action big model is a professional ability model, and the two have different ability characteristics and different application fields, which complement each other and complement the scene. LYDIA can realize the cognitive understanding of action data in the field of spatial computing, so as to generate actions efficiently. At the same time, LYDIA support is connected and invoked by other software platforms, so that the action files it generates can be used commercially by ecological partners. "

At the release site, Cui Chao, Technical Director of Liad Virtual Motion Point, first demonstrated the cognitive understanding and description ability of LYDIA large model to action data files, which can greatly improve the efficiency of AI data tagging industry and provide guarantee for large model continuous training of massive data. Secondly, the scene uses LYDIA large model to generate a simple action, a group of complex actions and a series of characteristic actions. Finally, the data connection and transfer ability between the LYDIA large model and the mainstream rendering engine Unity platform is demonstrated on the spot, and the commercial closed loop between the action large model and other tool platforms is realized.

LYDIA and OptiTrack

Serve different scenarios according to different needs

LYDIA and OptiTrack are two kinds of technical products with different capabilities. The large LYDIA model focuses on solving the industry scenarios where the efficiency of motion data acquisition is more important, while OptiTrack optical dynamic capture is better at the industry scenarios where precision is more important. In short, LYDIA pays attention to efficiency and Optitrack pays attention to precision. The two cooperate with each other and operate efficiently to create value for customers with different needs.

For example, in the film and television, e-commerce, animation, games and other industries, LYDIA can greatly improve the industry production efficiency and create greater value for the industry by virtue of its AIGC generation ability.

Bow into the game and forge ahead. In the context of the AI era, AIGC has become the core force of industry change. In the future, Liad Virtual moving Point will continue to improve the technical capability of LYDIA and join hands with many partners to build a "professional ability action model ecology".

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.