This dataset is a copy of a subset of the full Stanford Dogs Dataset.

Source:
http://vision.stanford.edu/aditya86/ImageNetDogs/

The original dataset contained 20,580 images of 120 breeds of dogs.

This subset contains 9884 images of 60 breeds of dogs.

Dataset Information

This dataset contains 14,674 images (12,444 of which contain objects of interest with bounding box annotations) of fish, crabs, and other marine animals. It was collected with a camera mounted 9 meters below the surface on the Limfjords bridge in northern Denmark by Aalborg University.

Composition

Roboflow has extracted and processed the frames from the source videos and converted the annotations for use with many popular computer vision models. We have maintained the same 80/10/10 train/valid/test split as the original dataset.

The class balance in the annotations is as follows:

Most of the identified objects are congregated towards the bottom of the frames.

More Information

For more information, see the Detection of Marine Animals in a New Underwater Dataset with Varying Visibility paper.

If you find the dataset useful, the authors request that you please cite their paper:

@InProceedings{pedersen2019brackish,
                                title={Detection of Marine Animals in a New Underwater Dataset with Varying Visibility},
                                author={Pedersen, Malte and Haurum, Joakim Bruslund and Gade, Rikke and Moeslund, Thomas B. and Madsen, Niels},
                                booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
                                month = {June},
                                year = {2019}
                            }
                            

Dataset Details

This dataset consists of 638 images collected by Roboflow from two aquariums in the United States: The Henry Doorly Zoo in Omaha (October 16, 2020) and the National Aquarium in Baltimore (November 14, 2020). The images were labeled for object detection by the Roboflow team (with some help from SageMaker Ground Truth). Images and annotations are released under a Creative Commons By-Attribution license. You are free to use them for any purposes personal, commercial, or academic provided you give acknowledgement of their source.

Projects Using this Dataset:

No-Code Object Detection Tutorial

Class Breakdown

The following classes are labeled: fish, jellyfish, penguins, sharks, puffins, stingrays, and starfish. Most images contain multiple bounding boxes.

Usage

The dataset is provided in many popular formats for easily training machine learning models. We have trained a model with CreateML (see gif above).

This dataset could be used for coral reef conservation, environmental health monitoring, swimmer safety, pet analytics, automated feeding, and much more. We're excited to see what you build!

Overview

This is an object detection dataset of ocean fish classified by their latin names.

Use Cases

This dataset can be used for the following purposes:

• Underwater object detection model
• Fish object detection model
• Train object detection model to recognize underwater species
• Prototype fish detection system
• Identifying fish with computer vision
• Free fish dataset
• Free fish identificaiton dataset
• Scuba diving object detection dataset
• Fish bounding boxes
• Fish species annotations

Enjoy! These images have been listed in the public domain.

Note: These images have been sourced from makeml.app/datasets/fish

BIRDSAI: A Dataset for Detection and Tracking in Aerial Thermal Infrared Videos

Authors:

• Elizabeth Bondi, Harvard University
• Raghav Jain, University of Southern California
• Palash Aggrawal, Indraprastha Institute of Information Technology
• Saket Anand, Indraprastha Institute of Information Technology
• Robert Hannaford, Duke University
• Ashish Kapoor, University of Delhi
• Jim Piavis, The Citadel
• Shital Shah, University of Mumbai
• Lucas Joppa, Chief Environmental Officer, Microsoft
• Bistra Dilkina, University of Southern California
• Milind Tambe, Harvard University

Published: 2020

Description: The Benchmarking IR Dataset for Surveillance with Aerial Intelligence (BIRDSAI, pronounced bird's-eye) is a long-wave thermal infrared dataset containing nighttime images of animals and humans in Southern Africa. The dataset allows for benchmarking of algorithms for automatic detection and tracking of humans and animals with both real and synthetic videos.

Use Cases: Wildlife Poaching Prevention, Night-time Intruder Detection, Wildlife Monitoring, Animal Behavior Research, Long Distance IR Detection

Download: The data can be downloaded from the Labeled Information Library of Alexandria

Training Dataset Download: https://lilablobssc.blob.core.windows.net/conservationdrones/v01/conservation_drones_train_real.zip

Annotation Format:
We follow the MOT annotation format, which is a CSV with the following columns:

<frame_number>, <object_id>, <x>, <y>, <w>, <h>, <class>, <species>, <occlusion>, <noise>

class: 0 if animals, 1 if humans

species: between -1 and 8 representing species below; 3 and 4 occur only in real data; 5, 6, 7, 8 occur only in synthetic data (note: most very small objects have unknown species)

-1: unknown, 0: human, 1: elephant, 2: lion, 3: giraffe, 4: dog, 5: crocodile, 6: hippo, 7: zebra, 8: rhino

occlusion: 0 if there is no occlusion, 1 if there is an occlusion (i.e., either occluding or occluded) (note: intersection over union threshold of 0.3 used to assign occlusion; more details in paper)

noise: 0 if there is no noise, 1 if there is noise (note: noise labels were interpolated from object locations in previous and next frames; for more than 4 consecutive frames without labels, no noise labels were included; more details in paper)

Acknowledgements: BIRDSAI was supported by Microsoft AI for Earth, NSF CCF-1522054 and IIS-1850477, MURI W911NF-17-1-0370, and the Infosys Center for Artificial Intelligence, IIIT-Delhi . Thanks to the labeling team and the Labeled Information Library of Alexandria for hosting the data.

Citation:

@inproceedings{bondi2020birdsai,
title={BIRDSAI: A Dataset for Detection and Tracking in Aerial Thermal Infrared Videos},
author={Bondi, Elizabeth and Jain, Raghav and Aggrawal, Palash and Anand, Saket and Hannaford, Robert and Kapoor, Ashish and Piavis, Jim and Shah, Shital and Joppa, Lucas and Dilkina, Bistra and Tambe, Milind},
booktitle={WACV},
year={2020}
}

• Elephants Dataset batch images cloned from:
  • https://universe.roboflow.com/eletrain/ele-dahd9
  • https://universe.roboflow.com/ttspl/elephant_dect
  • https://universe.roboflow.com/himanshi-shrivastava-p7p8v/ele_2
  • https://universe.roboflow.com/bitsathy/elephant_detection-ssczr
  • https://universe.roboflow.com/shaf-sekz/african-elephants/
  • https://universe.roboflow.com/popular-benchmarks/mit-indoor-scene-recognition/ (null images)
  • https://universe.roboflow.com/mohamed-traore-2ekkp/face-detection-mik1i/ (null images)

The bird dataset can be used to detect birds from multiple angles and distances during different types of weather and seasons.

Use the bird dataset and detection api to create computer vision applications for birding, bird feeding, bird counting, bird population health, seasonality of bird migrations, and more!

Example bird detection project: https://twitter.com/bradfordgill_/status/1509376362473209871?s=21&t=Ix8RrjaImfrKlJNi5if8iw

Use your home security camera to create notifications when birds have arrived by using code from this animal detection project: https://blog.roboflow.com/rabbit-deterrence-system/

Research paper on animal detection: https://ieeexplore.ieee.org/abstract/document/9752203

This dataset contains annotated pictures of animals (like wild pigs and deer) from trail cameras in East Texas.

You can use this dataset and the detection API to create computer vision applications for hunting, monitoring animal population health, counting deer sightings, and more!

Automatically filter through hours of trail cam footage to find the times/frames when wild game is caught on camera.

DetoxifAI Animals/Plants & Species:

• Snakes

• Coachwip (Masticophis flagellum)
                              

• Western Diamonback Rattlesnake (Crotalus atrox)
                              

• Pacific Gophersnake (Pituophis catenifer)
                              

• California kingsnake (Lampropeltis getula californiae)
                              

• Western yellow-bellied racer (Coluber constrictor)
                              

• Ring-necked snake (Diadophis punctatus)
                              

• Garter Snake (Thamnophis)
                              

• Sharp-tailed snake (Contia tenuis)
                              

• Rubber Boa (Charina bottae)
                              

• Northern Pacific rattelsnake (Crotalus Oreganus)
                              

• Mushrooms

• Black Trumpet (Craterellus cornucopioides)
                              

• Western Cauliflower Mushroom (Sparassis radicata)
                              

• Blushing Morel (Morchella rufobrunnea)
                              

• Pacific Golden Chanterelle (Cantharellus formosus)
                              

• Manzanita Bolete (Leccinum manzanitae)
                              

• Salt-Loving Mushroom (Agaricus bernardii)
                              

• Death Cap (Amanita phalloides)
                              

• Magpie (Coprinopsis picacea)
                              

• Fly Agaric (Amanita muscaria)
                              

• Jack-o-Lantern (Omphalotus olearius)
                              

Purpose of the Project

This project started as a way to add real-time counts of bees with/without pollen entering my backyard beehive to append some additional information to a livestream of the hive, and to correlate behavior at the hive entrance to weather, temperature, etc. Since then, I've added additional training data not specific to my hive which accounts for classification of drones and queens in addition to bees/pollen bees. Currently the model generalizes reasonably well, but more training data is required.

Assessed Classes & Labeling Guidelines

• bees (either workers or foragers)
• bees carrying pollen
• drones
• queens

Labeling should cover the entire body of the bee, excluding the wings as per the following example:

For the class of bees carrying pollen, it is acceptable to extend the box to include the visible pollen packs to distinguish this from the bee class:

Sample Results

Sample video of backyard hive entrance with low to moderate level of activity:
https://www.youtube.com/watch?v=qZW5eYd0Yw8&t=2266s

Generic sample video of single bee:
https://www.youtube.com/watch?v=A1x6VA8TWCg

Latest YOLOv5 Weights Files

Latest weights files for use by others are posted on my github here: https://github.com/mattnudi/bee-detection

These files will be updated as more images are added to the dataset

This is a dataset of bumble bee images curated by the Spiesman Lab at Kansas State University

• Brian Spiesman
• Research works: https://scholar.google.com/citations?user=U8xKpvgAAAAJ&hl=en

Project Overview
Creating a model to detect deer when driving down the road

Classes
There are 3 categories used to create the class for each photo. Example: buck_front_standing

1. Deer type - buck, doe, fawn; help model in case it improves results to distinguish between having antlers or not, having spots or not, etc
2. View of deer - front, rear, side; not only to help the model identify a deer that looks different from different angles (2 legs from the front) but also in case long term there is usefulness in identifying that a deer is running towards the road or away from it
3. Activity - standing, walking, running, eating; again, both to help the model but also for "threat" assessment as you drive and need to understand the current state of the deer

Status/Timeline
Initial images loaded (c. 60) to experiment with. Hope to have a larger dataset by 2023

Contribution and Labeling Guidelines
Any and all are welcome! We especially need deer in settings around roads.

About this Dataset

This is a collection of images and video frames of cheetahs at the Omaha Henry Doorly Zoo taken in October, 2020. The capture device was a SEEK Thermal Compact XR connected to an iPhone 11 Pro. Video frames were sampled and labeled by hand with bounding boxes for object detection using Robofow.

Using this Dataset

We have provided the dataset for download under a creative commons by-attribution license. You may use this dataset in any project (including for commercial use) but must cite Roboflow as the source.

Example Use Cases

This dataset could be used for conservation of endangered species, cataloging animals with a trail camera, gathering statistics on wildlife behavior, or experimenting with other thermal and infrared imagery.

About Roboflow

Roboflow creates tools that make computer vision easy to use for any developer, even if you're not a machine learning expert. You can use it to organize, label, inspect, convert, and export your image datasets. And even to train and deploy computer vision models with no code required.

About this Dataset

This dataset was created by exporting the Oxford Pets dataset from Roboflow Universe, generating a version with Modify Classes to drop all of the classes for the labeled dog breeds and consolidating all cat breeds under the label, "cat." The bounding boxes were also modified to incude the entirety of the cats within the images, rather than only their faces/heads.

Oxford Pets

• The Oxford Pets dataset (also known as the "dogs vs cats" dataset) is a collection of images and annotations labeling various breeds of dogs and cats. There are approximately 100 examples of each of the 37 breeds. This dataset contains the object detection portion of the original dataset with bounding boxes around the animals' heads.

• Origin: This dataset was collected by the Visual Geometry Group (VGG) at the University of Oxford.

Overview

The Aerial Sheep dataset contains images taken from a birds-eye view with instances of sheep in them. Images do not differentiate between gender or breed of sheep, instead grouping them into a single class named "sheep".

Example Footage

See RIIS's sheep counter application for additional use case examples.
Link - https://riis.com/blog/counting-sheep-using-drones-and-ai/

About RIIS

https://riis.com/about/

About this Dataset

The Oxford Pets dataset (also known as the "dogs vs cats" dataset) is a collection of images and annotations labeling various breeds of dogs and cats. There are approximately 100 examples of each of the 37 breeds. This dataset contains the object detection portion of the original dataset with bounding boxes around the animals' heads.

Origin

This dataset was collected by the Visual Geometry Group (VGG) at the University of Oxford.

This dataset was originally created by Dane Sprsiter. To see the current project, which may have been updated since this version, please go here: https://universe.roboflow.com/dane-sprsiter/barnyard.

This dataset is part of RF100, an Intel-sponsored initiative to create a new object detection benchmark for model generalizability.

Access the RF100 Github repo: https://github.com/roboflow-ai/roboflow-100-benchmark

This dataset was originally created by Anonymous. To see the current project, which may have been updated since this version, please go here: https://universe.roboflow.com/nasca37/peixos3.

This dataset is part of RF100, an Intel-sponsored initiative to create a new object detection benchmark for model generalizability.

Access the RF100 Github repo: https://github.com/roboflow-ai/roboflow-100-benchmark

This dataset was originally created by My Game Pics. To see the current project, which may have been updated since this version, please go here: https://universe.roboflow.com/my-game-pics/my-game-pics.

This dataset is part of RF100, an Intel-sponsored initiative to create a new object detection benchmark for model generalizability.

Access the RF100 Github repo: https://github.com/roboflow-ai/roboflow-100-benchmark

This dataset was originally created by Nazmuj Shakib Diip, Afraim, Shiam Prodhan. To see the current project, which may have been updated since this version, please go here: https://universe.roboflow.com/commolybroken/dataset-z2vab.

This dataset is part of RF100, an Intel-sponsored initiative to create a new object detection benchmark for model generalizability.

Access the RF100 Github repo: https://github.com/roboflow-ai/roboflow-100-benchmark

This dataset was originally created by Omar Kapur, wwblodge, Ricardo Jenez, Justin Jeng, Jeffrey Day. To see the current project, which may have been updated since this version, please go here: https://universe.roboflow.com/omarkapur-berkeley-edu/livestalk.

This dataset is part of RF100, an Intel-sponsored initiative to create a new object detection benchmark for model generalizability.

Access the RF100 Github repo: https://github.com/roboflow-ai/roboflow-100-benchmark

The project is for automated processing of home video camera feeds. This dataset includes both daytime and nighttime (IR) images, typically from perspective of a typical camera.

I suggest splitting the dataset and training two models: one for daytime and the other for nighttime. The nighttime pictures have a single channenl while the daytime ones have three channels, this results in significantly different features being trained. I identify if the image has one or three channels using the following shell command: identify -colorspace HSL -verbose "$f" | egrep -q "(Channel 0: 1-bit|red: 1-bit)"

The images are full size, so different sized models can be created. I've been training at 608x608. It includes many null images which have in the past triggered a false positives.

The classes are simply the things of interest I've seen from my house. In general this is more useful than the standard yolo classes, such as Zebra. However, you may want to have bear or some other wildlife. I've found squirrels are too small for my cameras to reliable pickup and detect. The perspective and framing of content is quite different from typical stock photos, so I think it makes a lot of sense to train the model using only images from ipcams.

Ideally, I will make models available for the many different tools people are using for AI already, including:
Deepstack / BlueIris
MotionEye
Frigate

Cat Detector

The aim for this project was to build a cat detector that can accuratlet identify my two cats to provide a signal to a water fountain. The waterflow is customised to the specic cat with one cat prefering a faster flow and the other a slower. By identifying which cat is in the proximity the fountain can activate only when necessary and provide a customised experience for the feline.

Classes
The project includes two classes:
1)vifslan
2)bubba
The classes are thus two seperate cats that share alot of similarties but also idiosyncracies.

Current status
Working prototype