Project Overview:

The original goal was to use this model to monitor my rowing workouts and learn more about computer vision. To monitor the workouts, I needed the ability to identify the individual digits on the rowing machine. With the help of Roboflow's computer vision tools, such as assisted labeling, I was able to more quickly prepare, test, deploy and improve my YOLOv5 model. https://i.imgur.com/X1kHoEm.png" alt="Example Annotated Image from the Dataset">

https://i.imgur.com/uKRnFZc.png" alt="Inference on a Test Image using the rfWidget"> * How to Use the rfWidget

Roboflow's Upload API, which is suitable for uploading images, video, and annotations, worked great with a custom app I developed to modify the predictions from the deployed model, and export them in a format that could be uploaded to my workspace on Roboflow. * Uploading Annotations with the Upload API * Uploading Annotations with Roboflow's Python Package

What took me weeks to develop can now be done with the help of a single click utilize Roboflow Train, and the Upload API for Active Learning (dataset and model improvement). https://i.imgur.com/dsMo5VM.png" alt="Training Results - Roboflow FAST Model">

Dataset Classes:

• 1, 2, 3, 4, 5, 6, 7, 8, 9, 90 (class "90" is a stand-in for the digit, zero)

This dataset consits of 841 images. There are images from a different rowing machine and also from this repo. Some scenes are illuminated with sunlight. Others have been cropped to include only the LCD. Digits like 7, 8, and 9 are underrepresented.

For more information:

• A full tutorial for creating and deploying this model with YOLOv5: https://vision.philbrockman.com/

Overview

Via https://rpc-dataset.github.io: * This dataset enjoys the following characteristics: (1) It is by far the largest dataset in terms of both product image quantity and product categories. (2) It includes single-product images taken in a controlled environment and multi-product images taken by the checkout system. (3) It provides different levels of annotations for the checkout images. Comparing with the existing datasets, ours is closer to the realistic setting and can derive a variety of research problems.

Use Cases

This dataset could be used to create an automatic item counter or checkout system using computer vision with Roboflow's API, Python Package, or other deployment options, such as Web Browser, iOS device, or to an Edge Device: https://docs.roboflow.com/inference/hosted-api.

Using this Dataset

This dataset has been licensed on a CC BY 4.0 license. You can copy, redistribute, and modify the images as long as there is appropriate credit to the authors of the dataset.

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. https://i.imgur.com/WHFqYSJ.png" alt="https://roboflow.com">

This dataset contains images of cottontail rabbits, that you might commonly find in your back yard in North America.

https://i.imgur.com/YiQI5Xn.png%5B/img%5D" alt="Cottontail Rabbits">

As we all know, rabbits can be quite a nuisance to our gardens and vegetables. That's why this dataset was used to train an object detection model that automatically recognizes rabbits, issuing a sound to deter them away.

Hardware:

• A Raspberry Pi
• External Power Supply - https://www.amazon.com/MakerHawk-Raspberry-Uninterruptible-Management-Expansion/dp/B082CVWH3R/ref=sr_1_6?crid=3LJGHA055O4VL&dchild=1&keywords=battery+for+raspberry+pi&qid=1623698007&sprefix=battery+for+raspbe%2Caps%2C184&sr=8-6
• Camera for the Pi - https://www.amazon.com/Arducam-Megapixels-Sensor-OV5647-Raspberry/dp/B012V1HEP4/ref=sr_1_6?dchild=1&keywords=Raspberry+Pi+camera&qid=1624689746&sr=8-6
• Bluetooth Speaker - https://www.amazon.com/AUDIOVOX-SP881BL-Portable-Bluetooth-Rechargeable/dp/B07F8N6KJ9/ref=sr_1_4?crid=2363N4JZD3B18&dchild=1&keywords=canz+bluetooth+speaker&qid=1626056945&sprefix=CANZ+bluetoot%2Caps%2C173&sr=8-4

Software:

• Public Dataset - https://public.roboflow.com/object-detection/eastern-cottontail-rabbits
• Roboflow Annotate - https://docs.roboflow.com/annotate
• Roboflow Train - https://docs.roboflow.com/train
• Roboflow Inference API - https://docs.roboflow.com/inference

Code repository here - https://github.com/roboflow-ai/rabbit-deterrence

**Road Sign Detection: Project Overview **

The Road Sign Detection project aims to develop a robust and accurate machine learning model for detecting and classifying road signs in real-time, using advanced computer vision techniques. This project serves as a critical component in the development of autonomous driving systems, intelligent transportation, and driver-assistance technologies, enhancing road safety by reliably identifying road signs under diverse conditions.

**Project Objectives **

Detection and Classification: Detect the presence of road signs in images or video frames and classify them accurately according to specific sign categories. Real-Time Performance: Optimize the model to achieve real-time inference speeds suitable for deployment in systems where latency is critical, such as autonomous vehicles or traffic monitoring systems. Generalization Across Environments: Ensure high performance across varied lighting, weather, and geographical conditions by training on a diverse dataset of annotated road signs. Classes and Tags This project involves multiple classes of road signs, which may include, but are not limited to:

• Stop Signs
• Speed Limit Signs
• Yield Signs
• Pedestrian Crossings
• No Entry Signs
• Warning Signs Each class represents a distinct type of road sign with specific implications for vehicle behavior and road safety.

Data Collection and Annotation

Dataset Size: 739 annotated images. Data Annotation: Each image has been manually annotated to include precise bounding boxes around each road sign, ensuring high-quality training data. Data Diversity: The dataset includes images taken from various perspectives, in different lighting conditions, and at varying levels of image clarity to improve the model's robustness. Current Status and Timeline Data Collection and Annotation: Completed. Model Training: Ongoing, with initial results demonstrating promising accuracy in detecting and classifying road signs. Deployment: Plans are underway to deploy the model on edge devices, making it suitable for use in real-world applications where immediate response times are critical. Project Timeline: The project is set to complete the final stages of training and optimization within the next two months, with active testing and iterative improvements ongoing. External Resources Project on Roboflow Universe: View Project on Roboflow Universe Documentation and API Reference: Detailed documentation on the dataset structure, model training parameters, and deployment options can be accessed within the Roboflow workspace. Contribution and Labeling Guidelines Contributors are welcome to expand the dataset by labeling additional road sign images and diversifying annotations. To maintain consistency:

Labeling Standards: Use bounding boxes to tightly enclose each road sign, ensuring no extra space or missing parts. Quality Control: Annotated images should be reviewed for accuracy, clarity, and proper categorization according to the predefined class types. This Road Sign Detection project is publicly listed on Roboflow Universe, where users and collaborators can download, contribute to, or learn more about the dataset and model performance.

Overview

The Insect_Detect_detection dataset contains images of an artifical flower platform with different insects sitting on it or flying above it. All images were automatically recorded with the Insect Detect DIY camera trap, a hardware combination of the Luxonis OAK-1, Raspberry Pi Zero 2 W and PiJuice Zero pHAT for automated insect monitoring.

Classes

The following object classes were annotated in this dataset:

• wasp (mostly Vespula sp.)
• hbee (Apis mellifera)
• fly (mostly Brachycera)
• hovfly (various Syrphidae, e.g. Eupeodes corollae, Episyrphus balteatus, Scaeva pyrastri)
• other (all Arthropods with insufficient occurences, e.g. various Hymenoptera, true bugs, beetles)
• shadow (shadows of the recorded insects)

View the Health Check for more info on class balance.

Versions

Different dataset versions are available for export: - v7 insect_detect_320_1class - squashed to square (aspect ratio 1:1) - downscaled to 320x320 pixel - all classes merged into one class (insect) - use this version to train a YOLO insect detection model for the DIY camera trap - v4 insect_detect_416_1class - squashed to square (aspect ratio 1:1) - downscaled to 416x416 pixel - all classes merged into one class (insect) - slower inference speed compared to 320x320 px model input - v5 insect_detect_raw_4K - original images in 4K resolution (3840x2160 pixel) - v6 insect_detect_bbox_crop - contains the cropped bounding boxes and was used to generate the Insect_Detect_classification dataset

Deployment

You can use this dataset as starting point to train your own insect detection models. Take a look at the YOLO detection model training instructions for more information.

To deploy the YOLO object detection models for automated insect monitoring, check out the provided Python scripts, available in the insect-detect GitHub repo. More details about the processing pipeline can be found in the Insect Detect Docs.

License

This dataset is licensed under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0)

Citation

If you use this dataset, please cite our paper:

Sittinger M, Uhler J, Pink M, Herz A (2024) Insect detect: An open-source DIY camera trap for automated insect monitoring. PLoS ONE 19(4): e0295474. https://doi.org/10.1371/journal.pone.0295474