notebooks

Overview of notebooks

What it is

A collection of computer vision tutorials maintained by Roboflow. The repository contains 61 model tutorials, 4 tracker tutorials, and 23 computer vision skills notebooks. It covers state-of-the-art models such as YOLOv11, SAM 2, Florence-2, PaliGemma 2, Qwen2.5-VL, RF-DETR, and others. Tasks include object detection, segmentation, pose estimation, OCR, and data extraction.

Key features

• 61 model tutorials covering architectures from ResNet to RF-DETR, YOLO11, SAM 3, and Qwen3-VL
• 4 tracker tutorials using RF-DETR with ByteTrack, SORT, OC-SORT, and McByte
• 23 computer vision skills notebooks for tasks like auto-annotation, speed estimation, player tracking, and image retrieval
• Tutorials for fine-tuning models on custom datasets (object detection, instance segmentation, classification, pose estimation, oriented bounding boxes)
• Zero-shot object detection and segmentation tutorials using Grounding DINO, YOLO-World, YOLOE, and Gemini
• Notebooks for OCR with GLM-OCR and LaTeX OCR with PaliGemma2

Use cases

• Learning to fine-tune YOLO, SAM, DETR, and transformer-based models on custom datasets
• Implementing object tracking in video streams with various trackers
• Automating dataset annotation with GroundedSAM and Grounding DINO
• Performing OCR and structured data extraction from images
• Benchmarking object detection models on the RF100 dataset
• Building sports analytics (basketball, football) and vehicle speed estimation systems

Getting started

See the repository README for instructions on opening notebooks in Colab, Kaggle, or SageMaker Studio Lab.

License and pricing

The project is open source. The specific license is not stated in the provided metadata.

Based on the project's public README and repository metadata.