NGC | Catalog
CatalogModelsMONAI Endoscopic Tool Segmentation

MONAI Endoscopic Tool Segmentation

For downloads and more information, please view on a desktop device.
Logo for MONAI Endoscopic Tool Segmentation


A pre-trained model for the endoscopic tool segmentation task.



Latest Version



April 4, 2023


88.5 MB

Model Overview

A pre-trained model for the endoscopic tool segmentation task and is trained using a flexible unet structure with an efficient-b2 [1] as the backbone and a UNet architecture [2] as the decoder. Datasets use private samples from Activ Surgical.

The PyTorch model and torchscript model are shared in google drive. Details can be found in large_files.yml file. Modify the "bundle_root" parameter specified in configs/train.json and configs/inference.json to reflect where models are downloaded. Expected directory path to place downloaded models is "models/" under "bundle_root".



Datasets used in this work were provided by Activ Surgical.

Since datasets are private, existing public datasets like EndoVis 2017 can be used to train a similar model.


When using EndoVis or any other dataset, it should be divided into "train", "valid" and "test" folders. Samples in each folder would better be images and converted to jpg format. Otherwise, "images", "labels", "val_images" and "val_labels" parameters in "configs/train.json" and "datalist" in "configs/inference.json" should be modified to fit given dataset. After that, "dataset_dir" parameter in "configs/train.json" and "configs/inference.json" should be changed to root folder which contains previous "train", "valid" and "test" folders.

Please notice that loading data operation in this bundle is adaptive. If images and labels are not in the same format, it may lead to a mismatching problem. For example, if images are in jpg format and labels are in npy format, PIL and Numpy readers will be used separately to load images and labels. Since these two readers have their own way to parse file's shape, loaded labels will be transpose of the correct ones and incur a missmatching problem.

Training configuration

The training as performed with the following:

  • GPU: At least 12GB of GPU memory
  • Actual Model Input: 736 x 480 x 3
  • Optimizer: Adam
  • Learning Rate: 1e-4


A three channel video frame


Two channels:

  • Label 1: tools
  • Label 0: everything else


IoU was used for evaluating the performance of the model. This model achieves a mean IoU score of 0.87.

Training Loss

A graph showing the training loss over 100 epochs.

Validation IoU

A graph showing the validation mean IoU over 100 epochs.

MONAI Bundle Commands

In addition to the Pythonic APIs, a few command line interfaces (CLI) are provided to interact with the bundle. The CLI supports flexible use cases, such as overriding configs at runtime and predefining arguments in a file.

For more details usage instructions, visit the MONAI Bundle Configuration Page.

Execute training:
python -m monai.bundle run training --meta_file configs/metadata.json --config_file configs/train.json --logging_file configs/logging.conf
Override the train config to execute evaluation with the trained model:
python -m monai.bundle run evaluating --meta_file configs/metadata.json --config_file "['configs/train.json','configs/evaluate.json']" --logging_file configs/logging.conf
Execute inference:
python -m monai.bundle run evaluating --meta_file configs/metadata.json --config_file configs/inference.json --logging_file configs/logging.conf
Export checkpoint to TorchScript file:
python -m monai.bundle ckpt_export network_def --filepath models/model.ts --ckpt_file models/ --meta_file configs/metadata.json --config_file configs/inference.json


[1] Tan, M. and Le, Q. V. Efficientnet: Rethinking model scaling for convolutional neural networks. ICML, 2019a.

[2] O. Ronneberger, P. Fischer, and T. Brox. U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical image computing and computer-assisted intervention, pages 234–241. Springer, 2015.


This training and inference pipeline was developed by NVIDIA. It is based on a model developed by NVIDIA researchers. This software has not been cleared or approved by FDA or any regulatory agency. MONAI pre-trained models are for developmental purposes only and cannot be used directly for clinical procedures.


Copyright (c) MONAI Consortium

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.