Tagged articles

large multimodal model

3 articles · Page 1 of 1

Jun 26, 2026 · Artificial Intelligence

From Human‑View Video to AI‑Understanding: Peking University’s Artic Framework Boosts Real‑Time AI Video Assistants

The Artic framework redesigns real‑time video communication for AI assistants by integrating model‑aware bitrate adaptation, region‑focused encoding, and a degradation‑aware benchmark, achieving a 15.12% accuracy gain and a 135.31 ms latency reduction in realistic mobile uplink scenarios while incurring modest cost overhead.

AI video communicationBenchmarkadaptive bitrate

0 likes · 11 min read

From Human‑View Video to AI‑Understanding: Peking University’s Artic Framework Boosts Real‑Time AI Video Assistants

AI Explorer

Apr 11, 2026 · Artificial Intelligence

Alibaba’s Wan2.7 Tops DesignArena: A Paradigm Shift for AI Video Creation

Alibaba’s Wan2.7 video model achieved a record‑high 1334 Elo score to win DesignArena, showcasing a leap in video understanding and generation that could reshape AI‑driven content creation, while also highlighting the massive compute demands, hallucination risks, and ethical challenges ahead.

AI Video UnderstandingAI market trendsDesignArena

0 likes · 5 min read

Alibaba’s Wan2.7 Tops DesignArena: A Paradigm Shift for AI Video Creation

Tencent Advertising Technology

Nov 22, 2022 · Artificial Intelligence

Tencent Advertising Multimedia AI: Systems, Technologies, and Business Applications

The article presents Tencent Advertising's comprehensive multimedia AI platform—including content understanding, intelligent creation, automated review, fingerprinting, and a large multimodal model—detailing their architectures, key techniques, and the significant performance and efficiency gains achieved across the ad ecosystem.

ad generationadvertising AIautomated review

0 likes · 18 min read

Tencent Advertising Multimedia AI: Systems, Technologies, and Business Applications