Skip to content
No description, website, or topics provided.
Python JavaScript Other
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.
eval_output numpy

You Only Look At CoefficienTs

    ██╗   ██╗ ██████╗ ██╗      █████╗  ██████╗████████╗
    ╚██╗ ██╔╝██╔═══██╗██║     ██╔══██╗██╔════╝╚══██╔══╝
     ╚████╔╝ ██║   ██║██║     ███████║██║        ██║   
      ╚██╔╝  ██║   ██║██║     ██╔══██║██║        ██║   
       ██║   ╚██████╔╝███████╗██║  ██║╚██████╗   ██║   
       ╚═╝    ╚═════╝ ╚══════╝╚═╝  ╚═╝ ╚═════╝   ╚═╝ 

A simple, fully convolutional model for real-time instance segmentation. This is the code for our paper, and for the forseeable future is still in development.

Here's a look at our current results for our base model (33 fps on a Titan Xp and 29.8 mAP on COCO's test-dev):

Example 0

Example 1

Example 2


  • Set up a Python3 environment.
  • Install Pytorch 1.0.1 (or higher) and TorchVision.
  • Install some other packages:
    # Cython needs to be installed before pycocotools
    pip install cython
    pip install opencv-python pillow pycocotools matplotlib 
  • Clone this repository and enter it:
    git clone
    cd yolact
  • If you'd like to train YOLACT, download the COCO dataset and the 2014/2017 annotations. Note that this script will take a while and dump 21gb of files into ./data/coco.
    sh data/scripts/
  • If you'd like to evaluate YOLACT on test-dev, download test-dev with this script.
    sh data/scripts/


As of April 5th, 2019 here are our latest models along with their FPS on a Titan Xp and mAP on test-dev:

Image Size Backbone FPS mAP Weights
550 Resnet50-FPN 42.5 28.2 yolact_resnet50_54_800000.pth Mirror
550 Darknet53-FPN 40.0 28.7 yolact_darknet53_54_800000.pth Mirror
550 Resnet101-FPN 33.0 29.8 yolact_base_54_800000.pth Mirror
700 Resnet101-FPN 23.6 31.2 yolact_im700_54_800000.pth Mirror

To evalute the model, put the corresponding weights file in the ./weights directory and run one of the following commands.

Quantitative Results on COCO

# Quantitatively evaluate a trained model on the entire validation set. Make sure you have COCO downloaded as above.
# This should get 29.92 validation mask mAP last time I checked.
python --trained_model=weights/yolact_base_54_800000.pth

# Output a COCOEval json to submit to the website or to use the script.
# This command will create './results/bbox_detections.json' and './results/mask_detections.json' for detection and instance segmentation respectively.
python --trained_model=weights/yolact_base_54_800000.pth --output_coco_json

# You can run COCOEval on the files created in the previous command. The performance should match my implementation in

# To output a coco json file for test-dev, make sure you have test-dev downloaded from above and go
python --trained_model=weights/yolact_base_54_800000.pth --output_coco_json --dataset=coco2017_testdev_dataset

Qualitative Results on COCO

# Display qualitative results on COCO. From here on I'll use a confidence threshold of 0.3.
python --trained_model=weights/yolact_base_54_800000.pth --score_threshold=0.3 --top_k=100 --display

Benchmarking on COCO

# Run just the raw model on the first 1k images of the validation set
python --trained_model=weights/yolact_base_54_800000.pth --benchmark --max_images=1000


# Display qualitative results on the specified image.
python --trained_model=weights/yolact_base_54_800000.pth --score_threshold=0.3 --top_k=100 --image=my_image.png

# Process an image and save it to another file.
python --trained_model=weights/yolact_base_54_800000.pth --score_threshold=0.3 --top_k=100 --image=input_image.png:output_image.png

# Process a whole folder of images.
python --trained_model=weights/yolact_base_54_800000.pth --score_threshold=0.3 --top_k=100 --images=path/to/input/folder:path/to/output/folder


# Display a video in real-time. "--video_multiframe" will process that many frames at once for improved performance.
python --trained_model=weights/yolact_base_54_800000.pth --score_threshold=0.3 --top_k=100 --video_multiframe=2 --video=my_video.mp4

# Display a webcam feed in real-time. If you have multiple webcams pass the index of the webcam you want instead of 0.
python --trained_model=weights/yolact_base_54_800000.pth --score_threshold=0.3 --top_k=100 --video_multiframe=2 --video=0

# Process a video and save it to another file. This is unoptimized.
python --trained_model=weights/yolact_base_54_800000.pth --score_threshold=0.3 --top_k=100 --video=input_video.mp4:output_video.mp4

As you can tell, can do a ton of stuff. Run the --help command to see everything it can do.

python --help


By default, we Train on COCO. Make sure to download the entire dataset using the commands above.

  • To train, grab an imagenet-pretrained model and put it in ./weights.
    • For Resnet101, download resnet101_reducedfc.pth from here.
    • For Resnet50, download resnet50-19c8e357.pth from here.
    • For Darknet53, download darknet53.pth from here.
  • Run one of the training commands below.
    • Note that you can press ctrl+c while training and it will save an *_interrupt.pth file at the current iteration.
    • All weights are saved in the ./weights directory by default with the file name <config>_<epoch>_<iter>.pth.
# Trains using the base config with a batch size of 8 (the default).
python --config=yolact_base_config

# Trains yolact_base_config with a batch_size of 5. For the 550px models, 1 batch takes up around 1.5 gigs of VRAM, so specify accordingly.
python --config=yolact_base_config --batch_size=5

# Resume training yolact_base with a specific weight file and start from the iteration specified in the weight file's name.
python --config=yolact_base_config --resume=weights/yolact_base_10_32100.pth --start_iter=-1

# Use the help option to see a description of all available command line arguments
python --help

Custom Datasets

You can also train on your own dataset by following these steps:

  • Create a COCO-style Object Detection JSON annotation file for your dataset. The specification for this can be found here. Note that we don't use some fields, so the following may be omitted:
    • info
    • liscense
    • Under image: license, flickr_url, coco_url, date_captured
    • categories (we use our own format for categories, see below)
  • Create a definition for your dataset under dataset_base in data/ (see the comments in dataset_base for an explanation of each field):
my_custom_dataset = dataset_base.copy({
    'name': 'My Dataset',

    'train_images': 'path_to_training_images',
    'train_info':   'path_to_training_annotation',

    'valid_images': 'path_to_validation_images',
    'valid_info':   'path_to_validation_annotation',

    'has_gt': True,
    'class_names': ('my_class_id_1', 'my_class_id_2', 'my_class_id_3', ...)
  • A couple things to note:
    • Class IDs in the annotation file should start at 1 and increase sequentially on the order of class_names. If this isn't the case for your annotation file (like in COCO), see the field label_map in dataset_base.
    • If you do not want to create a validation split, use the same image path and annotations file for validation. By default (see python --help), will output validation mAP for the first 5000 images in the dataset every 2 epochs.
  • Finally, in yolact_base_config in the same file, change the value for 'dataset' to 'my_custom_dataset' or whatever you named the config object above. Then you can use any of the training commands in the previous section.

Creating a Custom Dataset from Scratch

See this nice post by @Amit12690 for tips on how to annotate a custom dataset and prepare it for use with YOLACT.


If you use YOLACT or this code base in your work, please cite

  author    = {Daniel Bolya and Chong Zhou and Fanyi Xiao and Yong Jae Lee},
  title     = {YOLACT: {Real-time} Instance Segmentation},
  journal   = {arXiv},
  year      = {2019},


For questions about our paper or code, please contact Daniel Bolya.

You can’t perform that action at this time.