v0.6

Pre-built Linux binaries are available for the following environment:

CUDA	torch 1.10	torch 1.9	torch 1.8
11.3	install python -m pip install detectron2==0.6 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu113/torch1.10/index.html
11.1	install python -m pip install detectron2==0.6 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu111/torch1.10/index.html	install python -m pip install detectron2==0.6 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu111/torch1.9/index.html	install python -m pip install detectron2==0.6 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu111/torch1.8/index.html
10.2	install python -m pip install detectron2==0.6 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.10/index.html	install python -m pip install detectron2==0.6 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.9/index.html	install python -m pip install detectron2==0.6 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.8/index.html
10.1			install python -m pip install detectron2==0.6 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.8/index.html
cpu	install python -m pip install detectron2==0.6 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.10/index.html	install python -m pip install detectron2==0.6 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.9/index.html	install python -m pip install detectron2==0.6 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.8/index.html

v0.5

• Add “LazyConfig” system, an alternative python-based config system with more flexible syntax and no predefined structures.
• Add improved Mask R-CNN baselines to model zoo using large scale jittering and longer training. See blog and configs.
• Add backbone wrapper and detector wrapper for models in mmdetection. This allows training and evaluation of these models in detectron2. See an example config that trains a mmdet’s Mask R-CNN
• Code release for Implicit PointRend & PointSup.
• Code release for Rethinking Batch in BatchNorm.
• Add support for RegNet backbones. Example.
• Integration of fvcore’s tracing-based advanced flop counter.
• New features in DensePose CSE: see its release notes for more details.

Pre-built Linux binaries are available for the following environment:

CUDA	torch 1.9	torch 1.8	torch 1.7
11.1	install python -m pip install detectron2==0.5 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu111/torch1.9/index.html	install python -m pip install detectron2==0.5 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu111/torch1.8/index.html
11.0			install python -m pip install detectron2==0.5 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu110/torch1.7/index.html
10.2	install python -m pip install detectron2==0.5 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.9/index.html	install python -m pip install detectron2==0.5 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.8/index.html	install python -m pip install detectron2==0.5 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.7/index.html
10.1		install python -m pip install detectron2==0.5 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.8/index.html	install python -m pip install detectron2==0.5 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.7/index.html
9.2			install python -m pip install detectron2==0.5 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu92/torch1.7/index.html
cpu	install python -m pip install detectron2==0.5 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.9/index.html	install python -m pip install detectron2==0.5 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.8/index.html	install python -m pip install detectron2==0.5 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.7/index.html

v0.4

New Features

• All common models can be converted to TorchScript format by tracing or scripting (tutorial). Requires pytorch≥1.8.
• Support fvcore parameter schedulers (originally from ClassyVision) that are composable, scale-invariant, and can be used on parameters other than learning rate.
• Refactor PointRend as a mask head (instead of an ROIHead).
• New export and C++ deployment examples.
• Release d2go which provides end-to-end production pipeline.

New Features in DensePose:

Release DensePose CSE (a framework to extend DensePose to various categories using 3D models) and DensePose Evolution (a framework to bootstrap DensePose on unlabeled data). See here for more details.

Deprecations:

• Deprecate cfg argument from COCO/LVIS evaluator; Deprecate num_classes and ignore_label argument from SemSegEvaluator
• Deprecate WarmupMultiStepLR, WarmupCosineLR in favor of fvcore schedulers
• Deprecated features will be removed in future releases

Pre-built Linux binaries are available for the following environment:

CUDA	torch 1.8	torch 1.7	torch 1.6
11.1	install python -m pip install detectron2==0.4 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu111/torch1.8/index.html
11.0		install python -m pip install detectron2==0.4 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu110/torch1.7/index.html
10.2	install python -m pip install detectron2==0.4 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.8/index.html	install python -m pip install detectron2==0.4 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.7/index.html	install python -m pip install detectron2==0.4 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.6/index.html
10.1	install python -m pip install detectron2==0.4 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.8/index.html	install python -m pip install detectron2==0.4 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.7/index.html	install python -m pip install detectron2==0.4 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.6/index.html
9.2		install python -m pip install detectron2==0.4 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu92/torch1.7/index.html	install python -m pip install detectron2==0.4 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu92/torch1.6/index.html
cpu	install python -m pip install detectron2==0.4 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.8/index.html	install python -m pip install detectron2==0.4 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.7/index.html	install python -m pip install detectron2==0.4 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.6/index.html

v0.3

Features & Improvements:

• Support constructing RetinaNet, data loader, optimizer, COCOEvaluator without configs, in addition to Mask R-CNN.
• Add DeepLab & PanopticDeepLab in projects/.
• Support importing 3 projects (point_rend, deeplab, panoptic_deeplab) directly with import detectron2.projects.xxx.
• Support mixed precision in training (using cfg.SOLVER.AMP.ENABLED) and inference.
• Support ADE20k semantic segmentation dataset (named ade20k_sem_seg_train, ade20k_sem_seg_val).
• Continuous build on Windows.

Pre-built Linux binaries are provided for the following environment:

CUDA	torch 1.7	torch 1.6	torch 1.5
11.0	install python -m pip install detectron2==0.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu110/torch1.7/index.html
10.2	install python -m pip install detectron2==0.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.7/index.html	install python -m pip install detectron2==0.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.6/index.html	install python -m pip install detectron2==0.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.5/index.html
10.1	install python -m pip install detectron2==0.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.7/index.html	install python -m pip install detectron2==0.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.6/index.html	install python -m pip install detectron2==0.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.5/index.html
9.2	install python -m pip install detectron2==0.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu92/torch1.7/index.html	install python -m pip install detectron2==0.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu92/torch1.6/index.html	install python -m pip install detectron2==0.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu92/torch1.5/index.html
cpu	install python -m pip install detectron2==0.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.7/index.html	install python -m pip install detectron2==0.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.6/index.html	install python -m pip install detectron2==0.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.5/index.html

v0.2.1

• Added pre-built binary for PyTorch 1.6

CUDA	torch 1.6	torch 1.5	torch 1.4
10.2	install python -m pip install detectron2==0.2.1 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.6/index.html	install python -m pip install detectron2==0.2.1 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.5/index.html
10.1	install python -m pip install detectron2==0.2.1 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.6/index.html	install python -m pip install detectron2==0.2.1 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.5/index.html	install python -m pip install detectron2==0.2.1 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.4/index.html
10.0			install python -m pip install detectron2==0.2.1 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu100/torch1.4/index.html
9.2	install python -m pip install detectron2==0.2.1 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu92/torch1.6/index.html	install python -m pip install detectron2==0.2.1 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu92/torch1.5/index.html	install python -m pip install detectron2==0.2.1 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu92/torch1.4/index.html
cpu	install python -m pip install detectron2==0.2.1 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.6/index.html	install python -m pip install detectron2==0.2.1 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.5/index.html	install python -m pip install detectron2==0.2.1 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.4/index.html

v0.2

Features & Improvements:

• Support constructing objects with either configs or explicit arguments. As an example, the entire Mask R-CNN can be built without using configs
• Rename TransformGen to Augmentation and keep TransformGen as an alias. Design the interface of Augmentation so that it can access arbitrary custom data types. See augmentation tutorial for details.
• Improve speed of COCOEvaluator by about 3x
• Support LVIS v1 dataset
• Support GIoU loss in RPN and R-CNN
• Support auto-scaling of batch size and learning rate in DefaultTrainer. See cfg.SOLVER.REFERENCE_WORLD_SIZE

Pre-built Linux binaries are provided for the following environment:

CUDA	torch 1.5	torch 1.4
10.2	install python -m pip install detectron2==0.2 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.5/index.html
10.1	install python -m pip install detectron2==0.2 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.5/index.html	install python -m pip install detectron2==0.2 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.4/index.html
10.0		install python -m pip install detectron2==0.2 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu100/torch1.4/index.html
9.2	install python -m pip install detectron2==0.2 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu92/torch1.5/index.html	install python -m pip install detectron2==0.2 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu92/torch1.4/index.html
cpu	install python -m pip install detectron2==0.2 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.5/index.html	install python -m pip install detectron2==0.2 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.4/index.html

v0.1.3

Bugfix version.

We started to release pre-built wheels for multiple PyTorch versions:

CUDA	torch 1.5	torch 1.4
10.2	install python -m pip install detectron2==0.1.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu102/torch1.5/index.html
10.1	install python -m pip install detectron2==0.1.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.5/index.html	install python -m pip install detectron2==0.1.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.4/index.html
10.0		install python -m pip install detectron2==0.1.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu100/torch1.4/index.html
9.2	install python -m pip install detectron2==0.1.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu92/torch1.5/index.html	install python -m pip install detectron2==0.1.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cu92/torch1.4/index.html
cpu	install python -m pip install detectron2==0.1.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.5/index.html	install python -m pip install detectron2==0.1.3 -f \ https://dl.fbaipublicfiles.com/detectron2/wheels/cpu/torch1.4/index.html

Incompatible changes about internal interface:

• _init_{box,mask,keypoint}_head of StandardROIHeads was changed from instance method to class method.

v0.1.2

The pre-built wheels for this version have to be used with an official binary release of PyTorch 1.5.

Improvements:

• Add semantic segmentation models to PointRend
• Add examples to load a detectron2 model in C++
• New models that reproduce Rethinking ImageNet Pre-training
• Lots of new features in DensePose, see below
• Fix a few bugs in rotated box computation

Incompatible changes:

• When loading a checkpoint with resume_or_load(), training states like optimizer, start_iter will only be loaded when resume is True and the last checkpoint is found. This matches users’ expectations better
• .output_size in custom box head is renamed to .output_shape
• anchor_generator no longer duplicates the anchors for each image
• feature_strides and feature_channels attributes are removed from ROIHeads. Use the input argument input_shape instead.

New in DensePose:

• New evaluation metric (GPSm) that yields more reliable scores
• Panoptic FPN head implementation following Panoptic Feature Pyramid Networks
• DeepLabV3 head implementation following Rethinking Atrous Convolution for Semantic Image Segmentation
• Models with confidence estimation: implementation of the paper Correlated Uncertainty for Learning Dense Correspondences from Noisy Labels
• Contributions to the Model Zoo (for details please see the Model Zoo page):
  • Panoptic FPN and tuned hyperparameters for the original fully convolutional DensePose head: the performance of the existing R50 and R101 baselines improved by +10 AP
  • Panoptic FPN with DeepLabV3 head for DensePose: gives additional improvement of +2 AP
  • Models with confidence estimation
• Test time augmentations for DensePose (additional improvement of about +0.5 AP)

v0.1.1

Incompatible changes about head design:

• Mask head and keypoint head now include logic for losses & inference. Custom heads should overwrite the feature computation by layers() method.
• _forward_{box,mask,keypoint} methods of StandardROIHeads now accept dict of features.

This release is made to be compatible with such changes in projects (Mesh R-CNN, PointRend, etc)

Other additional features:

• flops & parameters counting
• Improve speed & RAM consumption of data loader

The pre-built wheels for this version have to be used with an official binary release of PyTorch 1.4.

Detectron2 v0.1 Release

Some major additional features since open source:

• Add TensorMask and PointRend in projects/.
• Support exporting models to caffe2 format.
• detectron2.model_zoo APIs.
• Support training with segmentation annotations in binary mask format.
• Support evaluating COCO-style AP for datasets in custom format.

We start to provide pre-built binary wheels at https://dl.fbaipublicfiles.com/detectron2/wheels/index.html.
The pre-built wheels for this version have to be used with an official binary release of PyTorch 1.4.