HOI-DETR is a transformer-based framework for detecting hands, hand-held objects, and their interactions in images and video. Built on the Co-DETR architecture, it adds a lightweight interaction ...

Buy thesis: premium GPU-as-a-service, $99.4B backlog, improving debt costs, and discounted valuation vs peers. See more here.

Llama 4 是 Meta 于 2025 年 4 月发布的多模态大语言模型系列，采用混合专家（MoE）架构，旗下包含 Scout（109B 总参数）、Maverick（400B 总参数）两个已开放权重的模型，以及仍在训练中的超旗舰 Behemoth（约 2T 总参数）。这一代模型原生支持图文多模态输入，最长 ...