Nvidia Unveils 'LocateAnything' AI Model for High-Speed, High-Accuracy Object Detection

Nvidia released LocateAnything, a 3B parameter open model on HuggingFace that outperforms Qwen3-VL and Rex-Omni on fine-grained detection. Released with CVPR202

📖 Read Full Pulse