NewCap: Xingpu Large Model Shows Excellent Performance, Significantly Reducing Computing Power Consumption

OpenAI Launches Pioneers Program to Redefine AI Model Evaluation

OpenAI has announced the launch of its 'OpenAI Pioneers Program', aimed at improving the current scoring system for AI models to create evaluation standards more aligned with real-world applications. With the rapid advancement of AI across various industries, understanding and enhancing AI's performance in real-world scenarios is crucial. OpenAI states that focusing on domain-specific evaluation metrics will more effectively reflect real-world performance and help teams assess model performance in high-stakes environments.

JD Retail Launches its First Self-Developed Billion-Parameter Time Series Model, TimeHF, for Predicting Product Sales

JD Retail's technology team announced the successful launch of TimeHF, its first self-developed billion-parameter time series model for sales forecasting. This model leverages Reinforcement Learning from Human Feedback (RLHF), applying it for the first time to sales forecasting. The accuracy of predictions has been significantly improved by over 10%, substantially reducing uncertainty in demand forecasting. This achievement has demonstrated excellent performance in JD's internal automated replenishment scenarios for 20,000 products.

Google Unveils AR Glasses Prototype: Seamlessly Blending Reality and the Digital World

At the latest TED conference, Google showcased a futuristic prototype of augmented reality (AR) glasses. The glasses, boasting a sleek design resembling regular eyewear, incorporate Google's advanced Gemini AI assistant, demonstrating impressive multi-functionality. In a demo, Shahram Izadi, head of the Android XR team, highlighted various applications, including real-time translation of Persian to English and book scanning. Izadi noted that the glasses...

Vision-R1: Reinforcing Visual Localization with RL, Achieving 50% Performance Boost

Recently, a collaborative effort between the Institute of Automation, Chinese Academy of Sciences, and the CASIA-Zidong Taichu team introduced Vision-R1, a novel method leveraging R1-like reinforcement learning to significantly enhance visual localization capabilities. This approach achieved a 50% performance improvement in complex tasks such as object detection and visual localization, surpassing existing state-of-the-art (SOTA) models with over ten times the parameter scale. Currently, large vision-language models typically rely on "pre-training + supervised fine-tuning" to improve responsiveness to user instructions, but this...

Sync Labs Releases Lipsync-2: The World's First Zero-Shot Lip-Sync Model

AI technology company Sync Labs recently announced via Twitter the launch of its latest product, Lipsync-2, touted as the world's first zero-shot lip-sync model. It retains the speaker's unique style without requiring additional training or fine-tuning. This breakthrough technology boasts significant improvements in realism, expressiveness, control, quality, and speed, suitable for real-person videos, animation, and AI-generated content. Lipsync-2's innovative features are based on Sync Lab's...

Quark AI-Powered! Alibaba's Smart AI Glasses Potentially Launching by the End of 2025

World's First Billion-Parameter AI Model Fine-tuning Task Successfully Runs on China's Quantum Computer, Origin Quantum's Wukong

A groundbreaking achievement in quantum computing has been reached with the successful fine-tuning of a billion-parameter AI model on Origin Quantum's Wukong, marking a significant milestone in the development of quantum-enhanced AI.

Guangxi High-Speed AI Traffic Model Deployed During Qingming Festival to Alleviate Congestion

During the Qingming Festival holiday, Guangxi province implemented an AI-powered traffic model on its highways to manage and alleviate traffic congestion.

Yuanding Intelligent Secures Nearly $100 Million in Funding, Partners with Fluidra to Reshape Industry Landscape

Yuanding Intelligent recently completed a strategic financing round of nearly 1 billion RMB (approximately $100 million USD). This round was led by Fluidra, a major player in the pool industry, with participation from new investor Yunqi Capital and existing investors. This financing marks a strategic turning point for Yuanding Intelligent, enabling it to penetrate global professional pool channels and transition from online leadership to full-channel penetration. Fluidra, with over 50 years of industry experience and operations spanning 47 countries, boasts well-known brands across Europe and America. Its global distribution network and other resources will support Yuanding Intelligent's transformation. Image Note: Image generated by AI, image licensing provided by Midjourney.

Wuhan's Physical Education Exam Integrates AI System, Benefiting Nearly 100,000 Students

From April 1st to 18th, approximately 100,000 ninth-grade students in Wuhan participated in their physical education entrance examination. This year's exam introduced a new AI intelligent judging system, marking a significant shift in physical examination assessment. Traditionally, physical education exams relied on manual timing and scoring. This year's exam utilized intelligent devices and cloud management to ensure fairness and efficiency. The exam incorporated several AI technologies, including facial recognition, electronic distance measurement, and wireless sensing, covering 14 items across three categories.