Roboflow

Best Keypoint Detection Models

Compare the best 3 keypoint detection models. 3 are open-weight, so you can self-host them for free under their licenses.

3 models · 3 open-weight · prices synced Aug 4, 2026

Open-Source Keypoint Detection Models

3 models with downloadable weights you can self-host under their licenses (Apache 2.0 and MIT).

Meta
Detectron2
Detectron2 is a computer vision model library developed by Facebook AI Research (Meta), released in September 2019. It serves as a comprehensive platform for object detection, instance segmentation, panoptic segmentation, keypoint detection, and DensePose, implemented in PyTorch. It is the successor to the original Detectron framework, which was written in Caffe2, and offers a more modular and extensible codebase designed for both research and production use.Detectron2 includes implementations of Faster R-CNN, Mask R-CNN, RetinaNet, Cascade R-CNN, Panoptic FPN, and several other architectures. Its modular design allows components such as backbones, necks, and heads to be swapped independently, making it widely used as a baseline framework in academic research. It supports training on COCO-format datasets and integrates with standard distributed training setups.
Google
MediaPipe
MediaPipe is an open-source framework developed by Google for building real-time machine learning pipelines across mobile, web, desktop, and edge platforms. First released in 2019, the framework uses a graph-based architecture where pre-built components called Calculators process streaming data such as images, video, and audio through configurable computation graphs. This design allows developers to compose perception pipelines from reusable building blocks without writing custom glue code between models. The current MediaPipe Tasks API replaces the earlier Solutions API and provides a unified cross-platform interface for vision, text, and audio.Rather than providing a single model, MediaPipe ships a suite of ready-to-use Tasks that wrap trained models for specific problems. These include MediaPipe Pose Landmarker for 33-point body landmark detection, Hand Landmarker for 21-point hand tracking, Face Landmarker which extends the earlier 468-point Face Mesh with blendshape outputs for facial expression, Selfie Segmentation for person-background separation, and Holistic Landmarker for combined body, hand, and face tracking. The Tasks prioritize on-device inference with low latency and support GPU acceleration where available, making the framework a common choice for mobile augmented reality, fitness and wellness applications, gesture-based interfaces, and accessibility features such as sign language recognition.
Meta
Mask R-CNN
Mask R-CNN is an instance segmentation model developed by Facebook AI Research (Meta), released in October 2017. It extends Faster R-CNN by adding a parallel branch that predicts binary segmentation masks for each detected object, independent of the classification and bounding box regression branches. A key contribution is RoIAlign, which replaces RoIPool with bilinear interpolation to preserve spatial correspondence between features and input pixels, significantly improving mask quality.Mask R-CNN achieves strong performance on the COCO instance segmentation benchmark and supports keypoint detection as an additional output head. It remains a foundational architecture in instance segmentation and is available through Meta's Detectron2 framework. The model is most appropriate for tasks requiring pixel-level object delineation, such as medical imaging, autonomous driving, and industrial inspection.

How to Choose a Keypoint Detection Model

Keypoint models localize precise points: joints, landmarks, corners. The choice depends on whether your keypoints are a standard set or custom to your objects.

Human body joints and face landmarks are solved categories with ready-made models (MediaPipe, keypoint heads in Detectron2 and Mask R-CNN): no training needed, well-understood accuracy, real-time options. If your points are on people or faces, start here.

Points specific to your objects (corners of a gauge, tips of an instrument, fiducials on a part) require a trained keypoint model on your own annotations. The labeling is cheap per image (a few clicks), and heatmap-based models localize with sub-pixel precision that boxes and masks cannot provide, which is exactly what alignment and measurement tasks need.

The bottom line: People and faces: use a pretrained keypoint model. Custom points on your own objects: train one; the annotation cost is low and nothing else gives you that localization precision.

Frequently Asked Questions About Keypoint Detection Models

Keypoint detection is the task of finding distinctive, precisely localized points in an image, such as facial landmarks, fingertips, joint centers, or the corners of an object. Models typically predict a heatmap per keypoint and take its peak, which yields sub-pixel localization that boxes and masks cannot provide. Human pose estimation is the best-known special case, but the same machinery aligns faces, registers medical images, anchors AR content, and tracks instrument tips. Evaluation compares predicted points to ground truth within a normalized distance (PCK-style metrics). This page lists 3 keypoint detection models, including 3 open-weight options you can self-host.

It depends on your task and constraints. For fixed categories in production, a model fine-tuned on your own data typically beats any general-purpose model. Compare the keypoint detection models on this page and try them on your own image to see which fits.

Yes. 3 of the 3 keypoint detection models here are open-weight (for example Detectron2, MediaPipe, and Mask R-CNN), free to self-host under their licenses (Apache 2.0 and MIT).

This page lists all 3 keypoint detection models in the Roboflow Playground catalog, all of them open-weight and free to self-host. Compare licenses, parameters, and release dates side by side, or open any model page for full details.