Search AI Products and News

Recently, global entertainment giants Disney and Universal Pictures have filed a copyright lawsuit against artificial intelligence company Midjourney, marking the first official legal dispute involving generative artificial intelligence for major Hollywood companies, drawing global attention. According to reports by CCTV, the two companies pointed out in their 110-page complaint that Midjourney allegedly illegally used its massive copyright library to generate and spread a large number of unauthorized copies of well-known characters, including characters from 'Star Wars,' 'Minions,' 'Shrek,' 'Frozen,' etc.

Pruna AI, a European startup focused on AI model compression algorithms, recently announced the open-sourcing of its optimization framework to help developers more efficiently compress AI models. The framework developed by Pruna AI combines several efficiency methods, including caching, pruning, quantization, and distillation, aiming to improve AI model performance. It not only standardizes the saving and loading of compressed models but also evaluates the compressed models to determine whether their quality has significantly degraded, while simultaneously measuring performance.

Chinese e-commerce giant Pinduoduo has announced the formation of a large model team focused on e-commerce recommendations, further strengthening its commitment to AI in the sector. According to recent posts on X, the team is led by a former core member of Baidu's Phoenix Nest (Baidu's advertising system), whose salary is reportedly multiple times higher than at Baidu, demonstrating Pinduoduo's significant investment in talent acquisition. While Pinduoduo has not officially disclosed the leader's identity, an X user @TechBitDaily revealed in a February 23rd post that the team has already recruited numerous individuals from Baidu.

In the field of artificial intelligence, the DeepSeek team recently released their latest research成果, introducing an innovative sparse attention mechanism named NSA (Native Sparse Attention). The core objective of this technology is to enhance the speed of long context training and inference, particularly optimized for modern hardware, significantly improving the efficiency of training and inference. The launch of NSA technology brings significant changes to the training of artificial intelligence models. Firstly, it utilizes a series of advancements tailored for modern computational hardware.

Google Research recently announced the innovative 'Titans' series model architecture, achieving a groundbreaking 2 million token context length through bionic design, with plans to open-source related technologies in the future. The core innovation of this architecture is the introduction of a Deep Neural Long-Term Memory Module, inspired by the human memory system. Titans cleverly combines the rapid response capability of short-term memory with the persistence characteristics of long-term memory, while utilizing an attention mechanism to handle immediate context, forming an efficient information processing system.

OpenAI recently announced the launch of its robot project to further explore the development of robotics technology. On January 11, Caitlin Kalinowski, the director of OpenAI's hardware department, revealed on social media that OpenAI would begin developing robots equipped with custom sensors and published related job openings on its official website. This recruitment covers three key positions: Electronic Perception Engineer, Robotics Mechanical Design Engineer, and Technical Project Manager. The primary responsibility of the Electronic Perception Engineer is to design and develop...

An innovative point cloud compression technology (TSC-PCAC) developed by a joint team from the Chinese Academy of Sciences, Tongji University, and Ningbo University has achieved a significant breakthrough. This technology not only greatly improves the compression efficiency of point cloud data but also significantly reduces processing time, clearing technical obstacles for the development of 3D applications like AR/VR. In the context of the rapid advancement of 3D vision technology, point clouds, as a key data format for virtual and augmented reality, face tremendous transmission and storage challenges. A high-quality point cloud can contain millions of data points, with each point...

Recently, the Meta AI team introduced LongVU, a novel spatio-temporal adaptive compression mechanism aimed at enhancing the language understanding capabilities of long videos. Traditional multimodal large language models (MLLMs) face limitations in context length when processing long videos, and LongVU was created to address this challenge. LongVU operates primarily by filtering duplicate frames and employing inter-frame token compression techniques to efficiently utilize context length, allowing it to reduce video data while preserving visual details.