Object Detection Models

YOLOv10

Tsinghua University|8M-68M

YOLOv10 is the tenth major iteration of the YOLO (You Only Look Once) real-time object detection series, developed by researchers at Tsinghua University. The model introduces a fundamentally redesigned NMS-free (Non-Maximum Suppression free) architecture that eliminates the post-processing bottleneck present in all previous YOLO versions, enabling true end-to-end object detection with consistent latency. YOLOv10 employs a dual-assignment training strategy that combines one-to-many and one-to-one label assignments during training, achieving rich supervision signals while maintaining efficient inference without redundant predictions. Built on a CSPNet backbone with enhanced feature aggregation, the model comes in six scale variants ranging from Nano (8M parameters) to Extra-Large (68M parameters), allowing deployment across edge devices, mobile platforms, and high-performance servers. Each variant is optimized for its target hardware profile, delivering the best accuracy-latency trade-off in its class. YOLOv10 achieves state-of-the-art performance on the COCO benchmark, outperforming previous YOLO versions and competing models like RT-DETR with significantly lower computational cost. Released under the AGPL-3.0 license, the model is open source and integrates seamlessly with the Ultralytics ecosystem for training, validation, and deployment. Common applications include autonomous driving perception, industrial quality inspection, security surveillance, retail analytics, robotics, and drone-based monitoring. The model supports ONNX and TensorRT export for optimized production deployment.

GroundingDINO

IDEA Research|172M

Grounding DINO is a powerful open-set object detection model developed by IDEA Research that locates and identifies any object in an image based on natural language text descriptions, representing a paradigm shift from fixed-category detection to language-guided visual understanding. With 172 million parameters, the model combines the DINO detection architecture with text grounding capabilities, enabling it to detect objects that were never seen during training simply by describing them in words. Unlike traditional object detectors trained on fixed categories like COCO's 80 classes, Grounding DINO can find arbitrary objects, parts, materials, or visual concepts by accepting free-form text queries such as 'red shoes on the shelf' or 'cracked window in the building.' The architecture fuses visual features from the image encoder with textual features from a text encoder through cross-modality attention layers, learning to align visual regions with their semantic descriptions. Grounding DINO achieves state-of-the-art results on zero-shot object detection benchmarks and when combined with SAM (Segment Anything Model) creates a powerful pipeline for text-prompted segmentation of any visual concept. Released under the Apache 2.0 license, the model is fully open source and widely used in computer vision research and production systems. Key applications include automated image annotation and labeling, visual search engines, robotic manipulation systems that understand verbal commands, visual question answering pipelines, content moderation systems, accessibility tools that describe image contents, and custom quality inspection systems that can be configured with natural language descriptions of defects rather than extensive training data.