# facebookresearch/map-anything

**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/facebookresearch-map-anything).**

2,915 stars · 207 forks · Python · apache-2.0

## Links

- GitHub: https://github.com/facebookresearch/map-anything
- awesome-repositories: https://awesome-repositories.com/repository/facebookresearch-map-anything.md

## Topics

`3d-reconstruction` `ai` `calibration` `depth-completion` `depth-estimation` `image-to-3d` `multi-view-stereo` `robotics` `sfm`

## Description

Map-anything is a 3D scene reconstruction framework and neural geometry estimator designed to transform two-dimensional images into metric three-dimensional spatial representations using feed-forward neural networks. It provides a specialized toolkit for predicting camera intrinsics and ray directions from single images without requiring external geometric metadata.

The project includes a 3D model benchmarking suite that utilizes a unified model wrapper to standardize outputs from diverse reconstruction models. This allows for consistent evaluation and accuracy measurement across various spatial datasets. To facilitate downstream use, it includes a COLMAP data exporter that converts neural reconstruction predictions into formats compatible with photogrammetry and splatting pipelines.

The framework covers a broad capability surface including distributed geometry model training, multi-node cluster orchestration, and inference memory optimization. It also provides tools for metric depth visualization, spatial data standardization, and geometry artifact filtering using normal-based masking.

## Tags

### Artificial Intelligence & ML

- [3D Reconstruction](https://awesome-repositories.com/f/artificial-intelligence-ml/foundation-models/3d-reconstruction.md) — Transforms two-dimensional images into metric three-dimensional spatial representations using feed-forward neural networks. ([source](https://github.com/facebookresearch/map-anything#readme))
- [3D Spatial AI](https://awesome-repositories.com/f/artificial-intelligence-ml/3d-spatial-ai.md) — Transforms two-dimensional images into three-dimensional spatial representations using a feed-forward metric network. ([source](https://github.com/facebookresearch/map-anything/blob/main/setup.py))
- [Metric Coordinate Mapping](https://awesome-repositories.com/f/artificial-intelligence-ml/feed-forward-neural-networks/metric-coordinate-mapping.md) — Transforms two-dimensional images into three-dimensional spatial representations using a neural network that predicts metric coordinates.
- [Model Benchmarking Suites](https://awesome-repositories.com/f/artificial-intelligence-ml/model-benchmarking-suites.md) — Implements a standardized interface for evaluating the accuracy of multiple 3D reconstruction models.
- [Geometry Model Training](https://awesome-repositories.com/f/artificial-intelligence-ml/model-training/pose-model-architectures/3d-pose-model-training/geometry-model-training.md) — Supports training and fine-tuning of reconstruction models using specialized pose loss functions. ([source](https://github.com/facebookresearch/map-anything#readme))
- [Reconstruction](https://awesome-repositories.com/f/artificial-intelligence-ml/text-to-sql-translators/accuracy-evaluators/reconstruction.md) — Evaluates reconstruction accuracy using standardized datasets across different numbers of input views. ([source](https://github.com/facebookresearch/map-anything/blob/main/benchmarking/dense_n_view/README.md))
- [Vision Dataset Standardizers](https://awesome-repositories.com/f/artificial-intelligence-ml/dataset-management/evaluation-datasets/evaluation-dataset-standardizers/vision-dataset-standardizers.md) — Converts diverse spatial datasets into a uniform format to streamline training and cross-model evaluation.
- [Image Data Preprocessing](https://awesome-repositories.com/f/artificial-intelligence-ml/image-data-preprocessing.md) — Performs undistortion and depth consistency calculations on image datasets to prepare them for 3D mapping. ([source](https://github.com/facebookresearch/map-anything/blob/main/data_processing/README.md))
- [Inference Memory Optimizations](https://awesome-repositories.com/f/artificial-intelligence-ml/inference-memory-optimizations.md) — Reduces GPU memory consumption during inference to allow processing of more views on limited hardware. ([source](https://github.com/facebookresearch/map-anything/blob/main/CHANGELOG.md))
- [Distributed Training](https://awesome-repositories.com/f/artificial-intelligence-ml/machine-learning/infrastructure/machine-learning-training/distributed-training.md) — Distributes large-scale geometry model training and fine-tuning workloads across multiple compute nodes. ([source](https://github.com/facebookresearch/map-anything/blob/main/train.md))
- [Reconstruction Output Standardization](https://awesome-repositories.com/f/artificial-intelligence-ml/model-output-formatting/reconstruction-output-standardization.md) — Wraps diverse reconstruction models in a uniform format to align 3D points, camera poses, and confidence scores. ([source](https://github.com/facebookresearch/map-anything/blob/main/README.md))
- [Third-Party Model Integration](https://awesome-repositories.com/f/artificial-intelligence-ml/third-party-model-integration.md) — Runs multiple third-party reconstruction models through a single interface to ensure consistent output formats for evaluation. ([source](https://github.com/facebookresearch/map-anything/blob/main/CHANGELOG.md))

### Part of an Awesome List

- [Metric 3D Scene Reconstruction](https://awesome-repositories.com/f/awesome-lists/ai/3d-reconstruction/metric-3d-scene-reconstruction.md) — Provides a comprehensive framework for transforming 2D images into metric 3D spatial representations.
- [3D Reconstruction Benchmarks](https://awesome-repositories.com/f/awesome-lists/learning/model-evaluation/3d-detection-benchmarks/3d-reconstruction-benchmarks.md) — Provides a benchmarking suite to standardize and measure the accuracy of diverse 3D reconstruction models.
- [Vision Model Fine-Tuning](https://awesome-repositories.com/f/awesome-lists/ai/model-training-and-fine-tuning/vision-model-fine-tuning.md) — Optimizes spatial reconstruction models using a modular training pipeline and comprehensive datasets. ([source](https://github.com/facebookresearch/map-anything/blob/main/train.md))
- [Photogrammetry Pipeline Integrations](https://awesome-repositories.com/f/awesome-lists/ai/photogrammetry-software/photogrammetry-pipeline-integrations.md) — Converts neural reconstruction predictions into formats compatible with photogrammetry tools like COLMAP.

### Graphics & Multimedia

- [Camera Intrinsic Predictions](https://awesome-repositories.com/f/graphics-multimedia/camera-intrinsic-predictions.md) — Recovers camera ray directions and intrinsic parameters from single images without requiring external geometric metadata. ([source](https://github.com/facebookresearch/map-anything/blob/main/benchmarking/calibration/README.md))
- [Ray-Direction Estimations](https://awesome-repositories.com/f/graphics-multimedia/ray-direction-estimations.md) — Recovers camera intrinsic parameters and ray directions from single images without requiring external geometric metadata.
- [Metric Depth Mapping](https://awesome-repositories.com/f/graphics-multimedia/depth-accuracy-metrics/metric-depth-mapping.md) — Processes depth maps and camera poses to maintain geometric consistency in 3D reconstructions.
- [Geometric Constraint Integrations](https://awesome-repositories.com/f/graphics-multimedia/geometric-constraint-integrations.md) — Integrates camera intrinsics, ray directions, and depth maps to improve the accuracy of 3D reconstructions. ([source](https://github.com/facebookresearch/map-anything#readme))
- [Geometry Artifact Filtering](https://awesome-repositories.com/f/graphics-multimedia/graphics-engines-rendering/3d-math-and-geometry-toolkits/geometry-primitives/coordinate-mapping/3d-geometry-utilities/geometry-artifact-filtering.md) — Denoises and removes edge artifacts from geometry outputs using normal-based masking and depth consistency checks. ([source](https://github.com/facebookresearch/map-anything/blob/main/README.md))
- [Metric 3D Representations](https://awesome-repositories.com/f/graphics-multimedia/metric-3d-representations.md) — Renders 3D reconstructed scenes using physically accurate spatial measurements to verify reconstruction quality.
- [Photogrammetry Format Exports](https://awesome-repositories.com/f/graphics-multimedia/photogrammetry-format-exports.md) — Converts reconstruction predictions into files compatible with external photogrammetry and splatting pipelines. ([source](https://github.com/facebookresearch/map-anything#readme))

### Scientific & Mathematical Computing

- [Camera Geometry Estimation](https://awesome-repositories.com/f/scientific-mathematical-computing/data-modeling-processing/geospatial-and-location-services/spatial-data-processing/spatial-geometry-libraries/camera-geometry-estimation.md) — Estimates camera ray directions and intrinsic parameters from single images without external metadata.
- [Geometry Masking](https://awesome-repositories.com/f/scientific-mathematical-computing/mask-based-filtering/geometry-masking.md) — Removes edge artifacts and low-confidence regions by applying depth consistency checks and normal-based filters to outputs.

### Data & Databases

- [Data Visualization](https://awesome-repositories.com/f/data-databases/data-analysis-visualization/visualization-frameworks-libraries/data-visualization.md) — Renders standardized 3D spatial data to verify the quality of neural reconstruction representations. ([source](https://github.com/facebookresearch/map-anything/tree/main/data_processing))
- [Data Export](https://awesome-repositories.com/f/data-databases/data-export.md) — Exports neural reconstruction predictions into structured formats compatible with photogrammetry and splatting pipelines.

### Software Engineering & Architecture

- [Unified Model Interfaces](https://awesome-repositories.com/f/software-engineering-architecture/unified-model-interfaces.md) — Wraps reconstruction models in a unified interface to ensure consistent output formats for benchmarking. ([source](https://github.com/facebookresearch/map-anything#readme))
- [Unified Model Wrappers](https://awesome-repositories.com/f/software-engineering-architecture/unified-model-wrappers.md) — Standardizes diverse third-party reconstruction models into a single interface to ensure consistent output formats for benchmarking.

### Testing & Quality Assurance

- [Reconstruction Benchmarking](https://awesome-repositories.com/f/testing-quality-assurance/reconstruction-benchmarking.md) — Executes standardized evaluation scripts using specific checkpoints and machine configurations to measure project performance. ([source](https://github.com/facebookresearch/map-anything/blob/main/benchmarking/rmvd_mvs_benchmark/README.md))