This Dataset contains HuBMap Dataset in COCO format to use in any Object Detection and Instance Segmentation Task.

COCO format easily supports Segmentation Frameworks such as AdelaiDet, Detectron2, TensorFlow etc.

The dataset is structured with images split into directories and no downscaling was done.

The following notebook explains how to convert custom annotations to COCO format:

https://www.kaggle.com/sreevishnudamodaran/build-custom-coco-annotations-512x512-tiled

Thanks to the Kaggle community and staff for all the support!

Please don't miss to upvote and comment if you like my work :)

Hope I everyone finds this useful!

Directory Structure:

   - coco_train
     - images(contains images in jpg format)
       - original_tiff_image_name
         - tile_column_number
           - image
           .
           .
           .
          .
          .
          .
        .
        .
        .
     - train.json (contains all the segmentation annotations in coco 
     -       format with proper relative path of the images)

Microsoft Common Objects in Context (COCO) Dataset

The Common Objects in Context (COCO) dataset is a widely recognized collection designed to spur object detection, segmentation, and captioning research. Created by Microsoft, COCO provides annotations, including object categories, keypoints, and more. The model it a valuable asset for machine learning practitioners and researchers. Today, many model architectures are benchmarked against COCO, which has enabled a standard system by which architectures can be compared.

While COCO is often touted to comprise over 300k images, it's pivotal to understand that this number includes diverse formats like keypoints, among others. Specifically, the labeled dataset for object detection stands at 123,272 images.

The full object detection labeled dataset is made available here, ensuring researchers have access to the most comprehensive data for their experiments. With that said, COCO has not released their test set annotations, meaning the test data doesn't come with labels. Thus, this data is not included in the dataset.

The Roboflow team has worked extensively with COCO. Here are a few links that may be helpful as you get started working with this dataset:

• An introduction to the COCO dataset
• Weird images in COCO, and what that tells us about the utility and limits of COCO

This dataset contains 1028 images each 640x380 pixels. The dataset is split into 249 test and 779 training examples. Every image comes with MS COCO format annotations. The dataset was collected in Carla Simulator, driving around in autopilot mode in various environments (Town01, Town02, Town03, Town04, Town05) and saving every i-th frame. The labels where then automatically generated using the semantic segmentation information.

Multitask COCO Dataset – Detection, Keypoints, and Segmentation

This publicly available Multitask COCO dataset has been preprocessed for seamless use in object detection, keypoint detection, and segmentation tasks. It enables multi-label annotations for COCO, ensuring robust performance across various vision applications. Special thanks to yermandy for providing access to multi-label annotations.

Optimized for deep learning models, this dataset is structured for easy integration into training pipelines, supporting diverse applications in computer vision research.

COCO is a large-scale object detection, segmentation, and captioning dataset.

Note: * Some images from the train and validation sets don't have annotations. * Coco 2014 and 2017 uses the same images, but different train/val/test splits * The test split don't have any annotations (only images). * Coco defines 91 classes but the data only uses 80 classes. * Panotptic annotations defines defines 200 classes but only uses 133.

To use this dataset:

import tensorflow_datasets as tfds

ds = tfds.load('coco', split='train')
for ex in ds.take(4):
 print(ex)

See the guide for more informations on tensorflow_datasets.

https://storage.googleapis.com/tfds-data/visualization/fig/coco-2014-1.1.0.png" alt="Visualization" width="500px">

DoPose (Dortmund Pose)is a dataset of highly cluttered and closely stacked objects. The dataset is saved in the BOP format. The dataset includes RGB images, Depth images, 6D Pose of objects, segmentation mask (all and visible), COCO Json annotation, camera transformations, and 3D model of all objects. The dataset contains 2 different types of scenes (table and bin). Each scene contains different view angles. For the bin scenes, the data contains 183 scenes with 2150 image views. In those 183 scenes 35 scenes contain 2 views, 20 contains 3 views and 128 contains 16 views. And for table scenes, the data contains 118 scenes with 1175 image views. in Those 118 scenes, 20 scenes contain 3 views, 50 scenes with 6 images, and 48 scenes with 17 images. So in total, our data contains 301 scenes and 3325 view images. Most of the scenes contain mixed objects. The dataset contains 19 objects in total.

For more info about the dataset content and collection process please refer to our Arxiv preprint

If you have any questions about the dataset, please contact anas.gouda@tu-dortmund.de

This is the full 2017 COCO object detection dataset (train and valid), which is a subset of the most recent 2020 COCO object detection dataset.

COCO is a large-scale object detection, segmentation, and captioning dataset of many object types easily recognizable by a 4-year-old. The data is initially collected and published by Microsoft. The original source of the data is here and the paper introducing the COCO dataset is here.

COCO (Common Objects in COntext) is a popular dataset in Computer Vision. It contains annotations for Computer Vision tasks - object detection, segmentation, keypoint detection, stuff segmentation, panoptic segmentation, densepose, and image captioning. For more details visit COCO Dataset

The Tensor Processing Unit (TPU) hardware accelerators are very fast. The challenge is often to feed them data fast enough to keep them busy. Google Cloud Storage (GCS) is capable of sustaining very high throughput but as with all cloud storage systems, initiating a connection costs some network back and forth. Therefore, having our data stored as thousands of individual files is not ideal. This dataset contains COCO dataset with object detection annotations in a smaller number of files and you can use the power of tf.data.Dataset to read from multiple files in parallel.

TFRecord file format Tensorflow's preferred file format for storing data is the protobuf-based TFRecord format. Other serialization formats would work too but you can load a dataset from TFRecord files directly by writing:

filenames = tf.io.gfile.glob(FILENAME_PATTERN) dataset = tf.data.TFRecordDataset(filenames) dataset = dataset.map(...)

For more details https://codelabs.developers.google.com/codelabs/keras-flowers-data/

You can use the following code in your kaggle notebook to get Google Cloud Storage (GCS) path of any public Kaggle dataset .

from kaggle_datasets import KaggleDatasets
GCS_PATH = KaggleDatasets().get_gcs_path()

View the notebook COCO Object Detection dataset in TFRecord to see how TFRecord files are created from the original COCO dataset.

Activities of Daily Living Object DatasetOverviewThe ADL (Activities of Daily Living) Object Dataset is a curated collection of images and annotations specifically focusing on objects commonly interacted with during daily living activities. This dataset is designed to facilitate research and development in assistive robotics in home environments.Data Sources and LicensingThe dataset comprises images and annotations sourced from four publicly available datasets:COCO DatasetLicense: Creative Commons Attribution 4.0 International (CC BY 4.0)License Link: https://creativecommons.org/licenses/by/4.0/Citation:Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., & Zitnick, C. L. (2014). Microsoft COCO: Common Objects in Context. European Conference on Computer Vision (ECCV), 740–755.Open Images DatasetLicense: Creative Commons Attribution 4.0 International (CC BY 4.0)License Link: https://creativecommons.org/licenses/by/4.0/Citation:Kuznetsova, A., Rom, H., Alldrin, N., Uijlings, J., Krasin, I., Pont-Tuset, J., Kamali, S., Popov, S., Malloci, M., Duerig, T., & Ferrari, V. (2020). The Open Images Dataset V6: Unified Image Classification, Object Detection, and Visual Relationship Detection at Scale. International Journal of Computer Vision, 128(7), 1956–1981.LVIS DatasetLicense: Creative Commons Attribution 4.0 International (CC BY 4.0)License Link: https://creativecommons.org/licenses/by/4.0/Citation:Gupta, A., Dollar, P., & Girshick, R. (2019). LVIS: A Dataset for Large Vocabulary Instance Segmentation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 5356–5364.Roboflow UniverseLicense: Creative Commons Attribution 4.0 International (CC BY 4.0)License Link: https://creativecommons.org/licenses/by/4.0/Citation: The following repositories from Roboflow Universe were used in compiling this dataset:Work, U. AI Based Automatic Stationery Billing System Data Dataset. 2022. Accessible at: https://universe.roboflow.com/university-work/ai-based-automatic-stationery-billing-system-data (accessed on 11 October 2024).Destruction, P.M. Pencilcase Dataset. 2023. Accessible at: https://universe.roboflow.com/project-mental-destruction/pencilcase-se7nb (accessed on 11 October 2024).Destruction, P.M. Final Project Dataset. 2023. Accessible at: https://universe.roboflow.com/project-mental-destruction/final-project-wsuvj (accessed on 11 October 2024).Personal. CSST106 Dataset. 2024. Accessible at: https://universe.roboflow.com/personal-pgkq6/csst106 (accessed on 11 October 2024).New-Workspace-kubz3. Pencilcase Dataset. 2022. Accessible at: https://universe.roboflow.com/new-workspace-kubz3/pencilcase-s9ag9 (accessed on 11 October 2024).Finespiralnotebook. Spiral Notebook Dataset. 2024. Accessible at: https://universe.roboflow.com/finespiralnotebook/spiral_notebook (accessed on 11 October 2024).Dairymilk. Classmate Dataset. 2024. Accessible at: https://universe.roboflow.com/dairymilk/classmate (accessed on 11 October 2024).Dziubatyi, M. Domace Zadanie Notebook Dataset. 2023. Accessible at: https://universe.roboflow.com/maksym-dziubatyi/domace-zadanie-notebook (accessed on 11 October 2024).One. Stationery Dataset. 2024. Accessible at: https://universe.roboflow.com/one-vrmjr/stationery-mxtt2 (accessed on 11 October 2024).jk001226. Liplip Dataset. 2024. Accessible at: https://universe.roboflow.com/jk001226/liplip (accessed on 11 October 2024).jk001226. Lip Dataset. 2024. Accessible at: https://universe.roboflow.com/jk001226/lip-uteep (accessed on 11 October 2024).Upwork5. Socks3 Dataset. 2022. Accessible at: https://universe.roboflow.com/upwork5/socks3 (accessed on 11 October 2024).Book. DeskTableLamps Material Dataset. 2024. Accessible at: https://universe.roboflow.com/book-mxasl/desktablelamps-material-rjbgd (accessed on 11 October 2024).Gary. Medicine Jar Dataset. 2024. Accessible at: https://universe.roboflow.com/gary-ofgwc/medicine-jar (accessed on 11 October 2024).TEST. Kolmarbnh Dataset. 2023. Accessible at: https://universe.roboflow.com/test-wj4qi/kolmarbnh (accessed on 11 October 2024).Tube. Tube Dataset. 2024. Accessible at: https://universe.roboflow.com/tube-nv2vt/tube-9ah9t (accessed on 11 October 2024). Staj. Canned Goods Dataset. 2024. Accessible at: https://universe.roboflow.com/staj-2ipmz/canned-goods-isxbi (accessed on 11 October 2024).Hussam, M. Wallet Dataset. 2024. Accessible at: https://universe.roboflow.com/mohamed-hussam-cq81o/wallet-sn9n2 (accessed on 14 October 2024).Training, K. Perfume Dataset. 2022. Accessible at: https://universe.roboflow.com/kdigital-training/perfume (accessed on 14 October 2024).Keyboards. Shoe-Walking Dataset. 2024. Accessible at: https://universe.roboflow.com/keyboards-tjtri/shoe-walking (accessed on 14 October 2024).MOMO. Toilet Paper Dataset. 2024. Accessible at: https://universe.roboflow.com/momo-nutwk/toilet-paper-wehrw (accessed on 14 October 2024).Project-zlrja. Toilet Paper Detection Dataset. 2024. Accessible at: https://universe.roboflow.com/project-zlrja/toilet-paper-detection (accessed on 14 October 2024).Govorkov, Y. Highlighter Detection Dataset. 2023. Accessible at: https://universe.roboflow.com/yuriy-govorkov-j9qrv/highlighter_detection (accessed on 14 October 2024).Stock. Plum Dataset. 2024. Accessible at: https://universe.roboflow.com/stock-qxdzf/plum-kdznw (accessed on 14 October 2024).Ibnu. Avocado Dataset. 2024. Accessible at: https://universe.roboflow.com/ibnu-h3cda/avocado-g9fsl (accessed on 14 October 2024).Molina, N. Detection Avocado Dataset. 2024. Accessible at: https://universe.roboflow.com/norberto-molina-zakki/detection-avocado (accessed on 14 October 2024).in Lab, V.F. Peach Dataset. 2023. Accessible at: https://universe.roboflow.com/vietnam-fruit-in-lab/peach-ejdry (accessed on 14 October 2024).Group, K. Tomato Detection 4 Dataset. 2023. Accessible at: https://universe.roboflow.com/kkabs-group-dkcni/tomato-detection-4 (accessed on 14 October 2024).Detection, M. Tomato Checker Dataset. 2024. Accessible at: https://universe.roboflow.com/money-detection-xez0r/tomato-checker (accessed on 14 October 2024).University, A.S. Smart Cam V1 Dataset. 2023. Accessible at: https://universe.roboflow.com/ain-shams-university-byja6/smart_cam_v1 (accessed on 14 October 2024).EMAD, S. Keysdetection Dataset. 2023. Accessible at: https://universe.roboflow.com/shehab-emad-n2q9i/keysdetection (accessed on 14 October 2024).Roads. Chips Dataset. 2024. Accessible at: https://universe.roboflow.com/roads-rvmaq/chips-a0us5 (accessed on 14 October 2024).workspace bgkzo, N. Object Dataset. 2021. Accessible at: https://universe.roboflow.com/new-workspace-bgkzo/object-eidim (accessed on 14 October 2024).Watch, W. Wrist Watch Dataset. 2024. Accessible at: https://universe.roboflow.com/wrist-watch/wrist-watch-0l25c (accessed on 14 October 2024).WYZUP. Milk Dataset. 2024. Accessible at: https://universe.roboflow.com/wyzup/milk-onbxt (accessed on 14 October 2024).AussieStuff. Food Dataset. 2024. Accessible at: https://universe.roboflow.com/aussiestuff/food-al9wr (accessed on 14 October 2024).Almukhametov, A. Pencils Color Dataset. 2023. Accessible at: https://universe.roboflow.com/almas-almukhametov-hs5jk/pencils-color (accessed on 14 October 2024).All images and annotations obtained from these datasets are released under the Creative Commons Attribution 4.0 International License (CC BY 4.0). This license permits sharing and adaptation of the material in any medium or format, for any purpose, even commercially, provided that appropriate credit is given, a link to the license is provided, and any changes made are indicated.Redistribution Permission:As all images and annotations are under the CC BY 4.0 license, we are legally permitted to redistribute this data within our dataset. We have complied with the license terms by:Providing appropriate attribution to the original creators.Including links to the CC BY 4.0 license.Indicating any changes made to the original material.Dataset StructureThe dataset includes:Images: High-quality images featuring ADL objects suitable for robotic manipulation.Annotations: Bounding boxes and class labels formatted in the YOLO (You Only Look Once) Darknet format.ClassesThe dataset focuses on objects commonly involved in daily living activities. A full list of object classes is provided in the classes.txt file.FormatImages: JPEG format.Annotations: Text files corresponding to each image, containing bounding box coordinates and class labels in YOLO Darknet format.How to Use the DatasetDownload the DatasetUnpack the Datasetunzip ADL_Object_Dataset.zipHow to Cite This DatasetIf you use this dataset in your research, please cite our paper:@article{shahria2024activities, title={Activities of Daily Living Object Dataset: Advancing Assistive Robotic Manipulation with a Tailored Dataset}, author={Shahria, Md Tanzil and Rahman, Mohammad H.}, journal={Sensors}, volume={24}, number={23}, pages={7566}, year={2024}, publisher={MDPI}}LicenseThis dataset is released under the Creative Commons Attribution 4.0 International License (CC BY 4.0).License Link: https://creativecommons.org/licenses/by/4.0/By using this dataset, you agree to provide appropriate credit, indicate if changes were made, and not impose additional restrictions beyond those of the original licenses.AcknowledgmentsWe gratefully acknowledge the use of data from the following open-source datasets, which were instrumental in the creation of our specialized ADL object dataset:COCO Dataset: We thank the creators and contributors of the COCO dataset for making their images and annotations publicly available under the CC BY 4.0 license.Open Images Dataset: We express our gratitude to the Open Images team for providing a comprehensive dataset of annotated images under the CC BY 4.0 license.LVIS Dataset: We appreciate the efforts of the LVIS dataset creators for releasing their extensive dataset under the CC BY 4.0 license.Roboflow Universe:

COCO 2017 Dataset for Image Colorization

  Overview

This dataset is derived from the COCO (Common Objects in Context) 2017 dataset, which is a large-scale object detection, segmentation, and captioning dataset. The COCO 2017 dataset has been adapted here specifically for the task of image colorization.

  Format

DatasetDict({ train: Dataset({ features: ['license', 'file_name', 'coco_url', 'height', 'width', 'date_captured', 'flickr_url', 'image_id'… See the full description on the dataset page: https://huggingface.co/datasets/nickpai/coco2017-colorization.

Dataset

This dataset was created by Sarkis Shil-Gevorkyan

Released under CC BY-SA 4.0

Contents

The IMPTOX project has received funding from the EU's H2020 framework programme for research and innovation under grant agreement n. 965173. Imptox is part of the European MNP cluster on human health.

More information about the project here.

Description: This repository includes the trained weights and a custom COCO-formatted dataset used for developing and testing a Faster R-CNN R_50_FPN_3x object detector, specifically designed to identify particles in micro-FTIR filter images.

Contents:

Weights File (neuralNetWeights_V3.pth):

Format: .pth

Description: This file contains the trained weights for a Faster R-CNN model with a ResNet-50 backbone and a Feature Pyramid Network (FPN), trained for 3x schedule. These weights are specifically tuned for detecting particles in micro-FTIR filter images.

Custom COCO Dataset (uFTIR_curated_square.v5-uftir_curated_square_2024-03-14.coco-segmentation.zip):

Format: .zip

Description: This zip archive contains a custom COCO-formatted dataset, including JPEG images and their corresponding annotation file. The dataset consists of images of micro-FTIR filters with annotated particles.

Contents:

Images: JPEG format images of micro-FTIR filters.

Annotations: A JSON file in COCO format providing detailed annotations of the particles in the images.

Management: The dataset can be managed and manipulated using the Pycocotools library, facilitating easy integration with existing COCO tools and workflows.

Applications: The provided weights and dataset are intended for researchers and practitioners in the field of microscopy and particle detection. The dataset and model can be used for further training, validation, and fine-tuning of object detection models in similar domains.

Usage Notes:

The neuralNetWeights_V3.pth file should be loaded into a PyTorch model compatible with the Faster R-CNN architecture, such as Detectron2.

The contents of uFTIR_curated_square.v5-uftir_curated_square_2024-03-14.coco-segmentation.zip should be extracted and can be used with any COCO-compatible object detection framework for training and evaluation purposes.

Code can be found on the related Github repository.

TACO is a growing image dataset of trash in the wild. It contains segmented images of litter taken under diverse environments: woods, roads and beaches. These images are manually labeled according to an hierarchical taxonomy to train and evaluate object detection algorithms. Annotations are provided in a similar format to COCO dataset.

For more information, go to: http://tacodataset.org

Dataset: COCO-Formatted Object Detection Dataset

Overview

This dataset is designed for object detection tasks and follows the COCO format. It contains 300 images and corresponding annotation files in JSON format. The dataset is split into training, validation, and test sets, ensuring a balanced distribution for model evaluation.

Dataset Structure

The dataset is organized into three main folders:

train/ (70% - 210 images)

valid/ (15% - 45 images)

test/ (15% - 45 images)

Each folder contains:

Images in JPEG/PNG format.

A corresponding _annotations.coco.json file that includes bounding box annotations.

Preprocessing & Augmentations

The dataset has undergone several preprocessing and augmentation steps to enhance model generalization:

Image Preprocessing:

Auto-orientation applied

Resized to 640x640 pixels (stretched)

Augmentation Techniques:

Flip: Horizontal flipping

Crop: 0% minimum zoom, 5% maximum zoom

Rotation: Between -5° and +5°

Saturation: Adjusted between -4% and +4%

Brightness: Adjusted between -10% and +10%

Blur: Up to 0px

Noise: Up to 0.1% of pixels

Bounding Box Augmentations:

Flipping, cropping, rotation, brightness adjustments, blur, and noise applied accordingly to maintain annotation consistency.

Annotation Format

The dataset follows the COCO (Common Objects in Context) format, which includes:

images section: Contains image metadata such as filename, width, and height.

annotations section: Includes bounding boxes, category IDs, and segmentation masks (if applicable).

categories section: Defines class labels.

Esefjorden Marine Vegetation Segmentation Dataset (EMVSD):Comprising 17,000 meticulously labeled images, this dataset is suited for instance segmentation tasks and represents a significant leap forward for marine research in the region. The images are stored in YOLO and COCO formats, ensuring compatibility with widely recognized and adopted object detection frameworks. Our decision to make this dataset publicly accessible underscores our commitment to collaborative research and the advancement of the broader scientific community.Dataset Structure:- Images: - Organized into three subsets: train, val, and test, located under the images/ directory. - Each subset contains high-resolution images optimized for object detection and segmentation tasks.- Annotations: - Available in YOLO txt and COCO formats for compatibility with major object detection frameworks. - Organized into three subsets: train, val, and test, located under the labels/ directory. - Additional metadata: - counts.txt: Summary of label distributions. - Cache files (train.cache, val.cache, test.cache) for efficient dataset loading.- Metadata: - classes.txt: Definitions for all annotated classes in the dataset. - Detailed COCO-format annotations in: - train_annotations.json - val_annotations.json - test_annotations.json- Configuration File: - EMVSD.yaml: Configuration file for seamless integration with machine learning libraries.Example Directory Structure:EMVSD/├── images/│ ├── train/│ ├── val/│ └── test/├── labels/│ ├── train/│ ├── val/│ ├── test/│ ├── counts.txt│ ├── train.cache│ ├── val.cache│ └── test.cache├── classes.txt├── train_annotations.json├── val_annotations.json├── test_annotations.json└── EMVSD.yaml

The Dataset

A collection of images of parking lots for vehicle detection, segmentation, and counting.
Each image is manually labeled with pixel-wise masks and bounding boxes localizing vehicle instances.
The dataset includes about 250 images depicting several parking areas describing most of the problematic situations that we can find in a real scenario: seven different cameras capture the images under various weather conditions and viewing angles. Another challenging aspect is the presence of partial occlusion patterns in many scenes such as obstacles (trees, lampposts, other cars) and shadowed cars.
The main peculiarity is that images are taken during the day and the night, showing utterly different lighting conditions.

We suggest a three-way split (train-validation-test). The train split contains images taken during the daytime while validation and test splits include images gathered at night.
In line with these splits we provide some annotation files:

• train_coco_annotations.json and val_coco_annotations.json --> JSON files that follow the golden standard MS COCO data format (for more info see https://cocodataset.org/#format-data) for the training and the validation splits, respectively. All the vehicles are labeled with the COCO category 'car'. They are suitable for vehicle detection and instance segmentation.

• train_dot_annotations.csv and val_dot_annotations.csv --> CSV files that contain xy coordinates of the centroids of the vehicles for the training and the validation splits, respectively. Dot annotation is commonly used for the visual counting task.

• ground_truth_test_counting.csv --> CSV file that contains the number of vehicles present in each image. It is only suitable for testing vehicle counting solutions.

Citing our work

If you found this dataset useful, please cite the following paper

@inproceedings{Ciampi_visapp_2021,
  doi = {10.5220/0010303401850195},
  url = {https://doi.org/10.5220%2F0010303401850195},
  year = 2021,
  publisher = {{SCITEPRESS} - Science and Technology Publications},
  author = {Luca Ciampi and Carlos Santiago and Joao Costeira and Claudio Gennaro and Giuseppe Amato},
  title = {Domain Adaptation for Traffic Density Estimation},
  booktitle = {Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications}
}

and this Zenodo Dataset

@dataset{ciampi_ndispark_6560823,
  author = {Luca Ciampi and Carlos Santiago and Joao Costeira and Claudio Gennaro and Giuseppe Amato},
  title = {{Night and Day Instance Segmented Park (NDISPark) Dataset: a Collection of Images taken by Day and by Night for Vehicle Detection, Segmentation and Counting in Parking Areas}},
  month = may,
  year = 2022,
  publisher = {Zenodo},
  version = {1.0.0},
  doi = {10.5281/zenodo.6560823},
  url = {https://doi.org/10.5281/zenodo.6560823}
}

Contact Information

If you would like further information about the dataset or if you experience any issues downloading files, please contact us at luca.ciampi@isti.cnr.it

SA-1B Download

Segment Anything 1 Billion (SA-1B) is a dataset designed for training general-purpose object segmentation models from open world images. The dataset was introduced in the paper "Segment Anything".

The SA-1B dataset consists of 11M diverse, high-resolution, licensed, and privacy-protecting images and 1.1B mask annotations. Masks are given in the COCO run-length encoding (RLE) format, and do not have classes.

The license is custom. Please, read the full terms and conditions on https://ai.facebook.com/datasets/segment-anything-downloads.

All the features are in the original dataset except image.content (content of the image).

You can decode segmentation masks with:

import tensorflow_datasets as tfds

pycocotools = tfds.core.lazy_imports.pycocotools

ds = tfds.load('segment_anything', split='train')
for example in tfds.as_numpy(ds):
 segmentation = example['annotations']['segmentation']
 for counts, size in zip(segmentation['counts'], segmentation['size']):
  encoded_mask = {'size': size, 'counts': counts}
  mask = pycocotools.decode(encoded_mask) # np.array(dtype=uint8) mask
  ...

To use this dataset:

import tensorflow_datasets as tfds

ds = tfds.load('segment_anything', split='train')
for ex in ds.take(4):
 print(ex)

See the guide for more informations on tensorflow_datasets.

Underwater Object Detection Dataset (COCO Format)

Dataset Overview

This dataset is structured for underwater object detection tasks, following the COCO annotation format. It contains both real and augmented images of various underwater objects (e.g., fish, coral, ROVs). Images are grouped into classes, and all annotations are stored in a single JSON file for ease of access and compatibility with most object detection frameworks.

Dataset Structure

The dataset folder structure is as follows:

Underwater_Object_Detection_Dataset/
├── combined_images/
│  ├── animal_fish/
│  │  ├── real_and_augmented_image1.jpg
│  │  ├── real_and_augmented_image2.jpg
│  │  └── ...
│  ├── plant/
│  │  ├── real_and_augmented_image1.jpg
│  │  └── ...
│  ├── rov/
│  │  ├── real_and_augmented_image1.jpg
│  │  └── ...
│  ├── test/
│  │  ├── test_image1.jpg
│  │  ├── test_image2.jpg
│  │  └── ...
│  ├── mixed_categories/
│  │  ├── mixed_image1.jpg
│  │  ├── mixed_image2.jpg
│  │  └── ...
│  └── ...
├── combined_annotations.json

Folder Details

• combined_images/: Contains subfolders for each class, with each folder containing both real and augmented images for that class.
• test/: Contains images specifically for testing the model, kept separate from the main classes.
• mixed_categories/: Contains images with multiple object classes in a single image, allowing for multi-object detection tasks.
• combined_annotations.json: A single JSON file with all image and annotation information, formatted in COCO-style for seamless integration with object detection models.

Annotations (combined_annotations.json)

The combined_annotations.json file follows the COCO format, structured into three main sections: images, annotations, and categories.

Example JSON Structure

{
  "images": [
    {
      "id": 1,
      "file_name": "vid_000159_frame0000008.jpg",
      "width": 480,
      "height": 270
    },
    {
      "id": 2,
      "file_name": "vid_000339_frame0000012.jpg",
      "width": 480,
      "height": 270
    }
    // Additional images
  ],
  
  "annotations": [
    {
      "segmentation": [],
      "area": 343.875,
      "iscrowd": 0,
      "image_id": 1,
      "bbox": [238.0, 165.0, 18.0, 23.0],
      "category_id": 1,
      "id": 221
    },
    {
      "segmentation": [],
      "area": 500.25,
      "iscrowd": 0,
      "image_id": 2,
      "bbox": [120.0, 140.0, 25.0, 20.0],
      "category_id": 2,
      "id": 222
    }
    // Additional annotations
  ],
  
  "categories": [
    {
      "supercategory": "marine_life",
      "id": 1,
      "name": "fish"
    },
    {
      "supercategory": "marine_life",
      "id": 2,
      "name": "coral"
    },
    {
      "supercategory": "vehicle",
      "id": 3,
      "name": "rov"
    }
    // Additional categories
  ]
}

JSON Key Explanations

• images: Contains metadata about each image:

  • "id": Unique identifier for the image.
  • "file_name": File name within its respective class folder.
  • "width" and "height": Dimensions of the image in pixels.

• annotations: Lists each object annotation with the following details:

  • "segmentation": For polygonal segmentation (empty here as we use bounding boxes only).
  • "area": Area of the bounding box.
  • "iscrowd": Set to 0 for individual objects, 1 if dense clustering.
  • "image_id": Corresponds to the id in images, linking the annotation to its image.
  • "bbox": Bounding box in [x_min, y_min, width, height] format.
  • "category_id": Refers to the object’s class in categories.
  • "id": Unique ID for each annotation.

• categories: Lists unique object classes in the dataset:

  • "supercategory": High-level grouping for the class.
  • "id": Unique ID for each class.
  • "name": Name of the object class.

Usage Recommendations

This dataset is suitable for: - Training and validation for underwater object detection models. - Benchmarking and testing on object detection algorithms. - Exploring domain adaptation using real and augmented underwater images.

Additional Notes

• The test/ folder is intended exclusively for testing the model, helping to evaluate its performance on unseen data.
• The mixed_categories/ folder contains images with multiple object types, making it suitable for multi-object detection challenges, where models need to detect several classes in the same image.

This dataset presents a comprehensive Bengali food segmentation dataset designed to support both semantic segmentation and object detection tasks using deep learning techniques. The dataset consists of high-quality images of traditional Bengali dishes captured in diverse real-life settings, annotated with polygon-based masks and categorized into multiple food classes. Annotation and preprocessing were performed using the Roboflow platform, with exports available in both COCO and mask formats. The dataset was used to train UNet for segmentation and YOLOv12 for detection. Augmentation and class balancing techniques were applied to improve model generalization. This dataset provides a valuable benchmark for food recognition, dietary assessment, and culturally contextualized computer vision research.

The DARRL dataset (Demonstrations for Action Recognition and Robot Learning) is a collection of 760 RGB-D videos of humans performing various manipulation tasks. It is provided with object and action annotations (in the COCO format) for 30 of those videos; segmentation masks are also provided.

It can also be used as a basis for learning from demonstrations for a robotic arm, for instance.

This work is supported by Région Pays de la Loire.