Collection of big model daily reports on January 11

News4months agorelease AIWindVane
20 0
Collection of big model daily reports on January 11

[Collection of big model daily reports on January 11] The ChatGPT application store is finally online, with more than 3 million GPTs in the entire network, and the era of sharing money with OpenAI is here; Shanghai AI Laboratory and other open source, audio and music unified development toolkit Amphion; the world’s first Full-stack self-developed full-scenario AI model for the automotive industry! Geely Xingrui AI large model officially released


Defeat 8 human planners: Tsinghua team proposes reinforcement learning urban space planning model

 

Link: https://news.miracleplus.com/share_link/15530

In recent years, in order to be more livable, the concept of “15-minute city” has received more attention. Its core is that residents can walk or ride to basic service facilities within 15 minutes, which reflects people’s desire for efficient spaces in urban communities. layout expectations. However, due to the diversity of urban geographical space, urban land layout and road planning is a very complex and difficult task, which has always relied heavily on the experience of professional planners. In response to this, the Urban Science and Computing Research Center of the Department of Electronics of Tsinghua University and the School of Architecture collaborated across disciplines to innovatively propose an urban community spatial planning method based on deep reinforcement learning. The proposed AI model can learn urban planning from massive data and continuously Optimize space efficiency and ultimately achieve planning levels beyond human experts.

Collection of big model daily reports on January 11

GauHuman open source: based on Gaussian Splatting, a high-quality 3D human body rapid reconstruction and real-time rendering framework

 

Link: https://news.miracleplus.com/share_link/15531

The Nanyang Technological University-SenseTime Joint Research Center S-Lab team proposed GauHuman, a high-quality 3D human body rapid reconstruction (1~2 minutes) and real-time rendering (up to 189 frames per second) framework based on Gaussian Splatting. GauHuman can quickly reconstruct the 3D human body based on a monocular human body video input by the user, as well as the corresponding camera and body motion body shape (SMPL) parameters, and render the 3D human body in real time.

Collection of big model daily reports on January 11

The ChatGPT application store is finally online. The number of GPTs on the entire network exceeds 3 million. The era of sharing money with OpenAI is here.

 

Link: https://news.miracleplus.com/share_link/15532

Will ChatGPT become a full-featured app like WeChat and DingTalk? On January 10, OpenAI’s application store GPT Store was officially launched, with categories, trends, weekly selections and other columns. Like most app stores, the GPT Store includes community rankings of popular downloads that users can search by categories such as writing, lifestyle, and education. OpenAI Greg Brockman said this is the first step in building your own ChatGPT. The product is still in a trial phase, but it is hoped to be rolled out more widely in the coming weeks.

Collection of big model daily reports on January 11

Many domestic universities have jointly built the open source community LAMM, and it’s time to join the multimodal language model family.

 

Link: https://news.miracleplus.com/share_link/15533

Scholars from Beihang University, Fudan University, University of Sydney, Chinese University of Hong Kong (Shenzhen) and Shanghai Artificial Intelligence Laboratory jointly launched one of the earliest open source communities for multi-modal language models – LAMM (Language-Assisted Multi-modal Model) . We aim to build LAMM into a growing community ecosystem to support research in MLLM training and evaluation, MLLM-driven Agents, and other directions. As one of the earliest open source projects in the field of multimodal large language models, LAMM’s goal is to establish an open research community ecosystem so that every researcher and developer can conduct research based on this and jointly build an open source community.

Collection of big model daily reports on January 11

Open source such as Shanghai AI Laboratory, audio and music unified development toolkit Amphion

 

Link: https://news.miracleplus.com/share_link/15534

The Shanghai AI Laboratory, the School of Data Science of the Chinese University of Hong Kong, and the Shenzhen Big Data Research Institute jointly open sourced an audio, music and speech generation toolkit called Amphion. Amphion can help developers study text-generated audio, music and other audio-related fields. It can be completed within a framework to solve problems such as black box generation models, scattered code bases, and lack of evaluation indicators. Amphion includes data processing, general modules, optimization algorithms and other infrastructure. At the same time, specific frameworks, models and development instructions are provided for tasks such as text-to-speech, singing voice conversion, and text-to-audio generation. It also has built-in various neural speech codecs and evaluation indicators. Especially for those new to generative AI development, Amphion is very easy to use.

Collection of big model daily reports on January 11

The world’s first full-stack self-developed full-scenario AI model for the automotive industry! Geely Xingrui AI large model officially released

 

Link: https://news.miracleplus.com/share_link/15535

After Gan Jiayue, CEO of Geely Automobile Group, released the relevant capabilities and multiple applications of AI large models at the Geely Galaxy E8 launch conference, on January 11, Geely officially released the world’s first full-stack self-developed full-scenario AI in the automotive industry. Large model – Geely Xingrui AI large model.

Collection of big model daily reports on January 11
© Copyright notes

Related posts

No comments

No comments...