# mikel-brostrom/boxmot **Attribution required: if you use, quote, or summarise this content, you must credit and link back to [awesome-repositories.com](https://awesome-repositories.com/repository/mikel-brostrom-boxmot).** 8,020 stars · 1,888 forks · Python · agpl-3.0 ## Links - GitHub: https://github.com/mikel-brostrom/boxmot - Homepage: https://deepwiki.com/mikel-brostrom/boxmot - awesome-repositories: https://awesome-repositories.com/repository/mikel-brostrom-boxmot.md ## Topics `boosttrack` `botsort` `bytetrack` `clip` `deep-learning` `deepocsort` `improvedassociation` `machine-learning` `mot` `mots` `multi-object-tracking` `multi-object-tracking-segmentation` `ocsort` `oriented-bounding-box-tracking` `osnet` `segmentation` `strongsort` `tensorrt` `tracking-by-detection` `yolo` ## Description Boxmot is a multi-object tracking framework designed to follow multiple objects across video frames using motion and appearance algorithms to maintain consistent identities. It functions as a system for tracking objects with specific orientations using rotated bounding boxes and corresponding intersection-over-union computations. The project includes a re-identification model optimizer that converts neural networks into formats for hardware-accelerated execution. It also features an evolutionary hyperparameter tuner that iteratively mutates tracker settings to maximize accuracy for specific datasets. The framework provides capabilities for computer vision benchmarking, including the use of persistent detection caching to speed up evaluation cycles. Additional functionality includes the ability to integrate tracking logic into native compiled applications via build tools. ## Tags ### Artificial Intelligence & ML - [Object Tracking Frameworks](https://awesome-repositories.com/f/artificial-intelligence-ml/object-tracking-frameworks.md) — Provides a framework for following multiple objects across video frames using motion and appearance algorithms. ([source](https://cdn.jsdelivr.net/gh/mikel-brostrom/boxmot@master/README.md)) - [Rotated](https://awesome-repositories.com/f/artificial-intelligence-ml/computer-vision-systems/computer-vision/object-detection-tracking/object-detection/rotated.md) — Supports tracking items with angled geometry using rotated bounding boxes to maintain accuracy. ([source](https://deepwiki.com/mikel-brostrom/boxmot)) - [Appearance Matching Networks](https://awesome-repositories.com/f/artificial-intelligence-ml/deep-learning-architectures/appearance-matching-networks.md) — Compares neural network embeddings of object visual traits across frames to resolve identities and handle occlusions. - [Appearance Embeddings](https://awesome-repositories.com/f/artificial-intelligence-ml/face-detection/identity-matching/appearance-embeddings.md) — Uses deep learning embeddings to compare object visual traits across frames and resolve identity switches. - [Rotated IoU Calculators](https://awesome-repositories.com/f/artificial-intelligence-ml/machine-learning/architectures/neural-network-components/loss-functions/perceptual-loss/content-loss-calculators/focal-loss-calculators/detection-loss-calculators/iou-aware-loss-calculators/rotated-iou-calculators.md) — Calculates intersection-over-union for oriented bounding boxes to maintain tracking accuracy for objects at varying angles. - [Rotated IoU Computations](https://awesome-repositories.com/f/artificial-intelligence-ml/machine-learning/frameworks/computer-vision/computer-vision-techniques/bounding-box-metrics/rotated-iou-computations.md) — Uses rotated geometry and specialized intersection-over-union computations to track objects with specific orientations. - [Rotated](https://awesome-repositories.com/f/artificial-intelligence-ml/object-tracking/rotated.md) — Implements tracking for objects with specific orientations using rotated bounding boxes to improve accuracy for angled items. - [Computer Vision Benchmarks](https://awesome-repositories.com/f/artificial-intelligence-ml/computer-vision-benchmarks.md) — Uses standardized evaluation suites to measure the accuracy and consistency of visual tracking systems. - [Computer Vision Evaluation Tools](https://awesome-repositories.com/f/artificial-intelligence-ml/computer-vision-evaluation-tools.md) — Ships a performance measurement system to evaluate tracking precision and identity consistency using standard datasets. - [Model Deployment](https://awesome-repositories.com/f/artificial-intelligence-ml/computer-vision-models/model-deployment.md) — Provides capabilities to export re-identification models into optimized formats for production inference on diverse hardware. - [Model Performance Benchmarks](https://awesome-repositories.com/f/artificial-intelligence-ml/cross-model-comparators/model-performance-benchmarks.md) — Allows comparing identification models by ranking them based on performance metrics using consistent detection sets. ([source](https://cdn.jsdelivr.net/gh/mikel-brostrom/boxmot@master/README.md)) - [Evolutionary Hyperparameter Tuners](https://awesome-repositories.com/f/artificial-intelligence-ml/evolutionary-algorithms/evolutionary-hyperparameter-tuners.md) — Provides an automated optimizer that uses evolutionary algorithms to find the most accurate tracker settings. - [Hyperparameter Optimization](https://awesome-repositories.com/f/artificial-intelligence-ml/hyperparameter-optimization.md) — Includes tools for searching and tuning tracker parameters using evolutionary algorithms. ([source](https://deepwiki.com/mikel-brostrom/boxmot)) - [ONNX Model Exporters](https://awesome-repositories.com/f/artificial-intelligence-ml/machine-learning/infrastructure/model-deployment-and-serving/serialization-and-export-formats/onnx-model-exporters.md) — Converts re-identification neural networks into the standardized ONNX format for hardware-accelerated execution. - [Re-Identification Model Optimizers](https://awesome-repositories.com/f/artificial-intelligence-ml/machine-learning/infrastructure/model-optimization-and-inference/serving-and-runtime/large-language-model-optimization/model-inference-optimizations/onnx-model-optimizers/re-identification-model-optimizers.md) — Provides a tool to convert re-identification neural networks into ONNX and TensorRT formats for hardware acceleration. - [Hyperparameter Tuning](https://awesome-repositories.com/f/artificial-intelligence-ml/machine-learning/infrastructure/model-optimization-and-inference/training-algorithms/machine-learning-optimization/hyperparameter-tuning.md) — Implements an iterative process for optimizing tracker hyperparameters to improve predictive accuracy. - [Deployment Optimizations](https://awesome-repositories.com/f/artificial-intelligence-ml/model-optimization/inference-deployment/deployment-optimizations.md) — Refines re-identification models for production execution by converting them to ONNX or TensorRT formats. - [Model Format Converters](https://awesome-repositories.com/f/artificial-intelligence-ml/neural-networks/model-format-converters.md) — Provides utilities to translate re-identification model weights and architectures into deployment formats. ([source](https://deepwiki.com/mikel-brostrom/boxmot)) ### Part of an Awesome List - [Evaluation and Benchmarking](https://awesome-repositories.com/f/awesome-lists/ai/evaluation-and-benchmarking.md) — Provides frameworks for measuring tracking precision and identity consistency using standard datasets. ([source](https://deepwiki.com/mikel-brostrom/boxmot)) - [Parameter Mutators](https://awesome-repositories.com/f/awesome-lists/ai/evolutionary-algorithms/parameter-mutators.md) — Iteratively mutates tracker hyperparameters to find the configuration that maximizes accuracy for a specific dataset.