Fire_Dataset

Real Fire Fire (283 Original images; 1,698 Augmented Images) Smoke (273 Original images; 1,638 Augmented Images)

No Fire Safe Fire (270 Original images; 1,620 Augmented Images) Artificial Fire (286 Original images; 1,716 Augmented Images)

Augmentation pipeline (including shear transformation using Affine) transform = A.Compose([ A.HorizontalFlip(p=0.5), # 50% chance to flip horizontally A.VerticalFlip(p=0.3), # 30% chance to flip vertically A.RandomBrightnessContrast(p=0.3), # Adjust brightness & contrast randomly A.Rotate(limit=30, p=0.5), # Rotate between -30 to +30 degrees A.Affine(shear=(-20, 20), p=0.5), # Apply shear transformation in both X & Y directions A.GaussNoise(var_limit=(10.0, 50.0), p=0.3) # Add random noise ])

The study, titled 'Effectiveness of Image Augmentation Techniques on Detection of Building Characteristics from Street View Images Using Deep Learning,' This paper investigates the impact of eight distinct image augmentation techniques—Brightness, Contrast, Perspective, Rotate, Scale, Shear, Translate, and a combined method termed 'A Sum of Techniques'—on two specific tasks: classifying building stories and identifying building typologies. The primary aim of this research is to enhance accuracy in these tasks using the aforementioned augmentation techniques. Additionally, this study emphasizes reproducibility of the proposed methods. All images used in the research were annotated specifically for the classification task.

🦴 Human Bone Fractures Multi-modal Image Dataset (HBFMID) 🦾

This dataset, the Human Bone Fractures Multi-modal Image Dataset (HBFMID), is a collection of medical images (X-ray and MRI) focused on detecting bone fractures in various parts of the human body. It's designed to support research in computer vision and deep learning for medical applications. 🧑‍⚕️💻

📄 Description

The HBFMID dataset contains a total of 1539 images of human bones, including both X-ray and MRI modalities. The dataset covers a wide range of bone locations, such as:

• Elbow
• Finger
• Forearm
• Humerus
• Shoulder
• Femur
• Shinbone
• Knee
• Hipbone
• Wrist
• Spinal cord
• Some healthy bones

The initial dataset consisted of 641 raw images (510 X-ray and 131 MRI). This raw data was then divided into three subsets:

• Training: 449 images
• Validation: 128 images
• Testing: 64 images

The images were carefully annotated to label the presence and location of fractures. ✍️

⚙️ Pre-processing

The following pre-processing steps were applied to the images:

• Auto-orientation: Ensuring consistent image orientation. 🔄
• Resizing: All images were resized to 640x640 pixels. 📏
• Contrast adjustments: Enhancing image clarity. ☀️

➕ Augmentation

To increase the dataset size and improve the robustness of machine learning models, various augmentation techniques were applied to the training set, resulting in approximately 1347 training images (449 x 3). The augmentation techniques included:

• Flip: Horizontal and Vertical (50% probability each) ↔️↕️
• Rotation: Random rotation between -5° and +5°. 🔄
• Shear: Random shear of ±2° horizontally and ±2° vertically. 📐
• Saturation: Random adjustment between -5% and +5%. 🌈
• Brightness: Random adjustment between -10% and +10%. 💡
• Zooming: 2%. 🔍

🛠️ Exported via Roboflow

This dataset was exported using Roboflow, an end-to-end computer vision platform that facilitates:

• Team collaboration on computer vision projects.
• Image collection and organization.
• Understanding and searching unstructured image data.
• Annotation and dataset creation.
• Model training and deployment.
• Active learning for continuous dataset improvement.

For state-of-the-art Computer Vision training notebooks compatible with this dataset, visit https://github.com/roboflow/notebooks. 🚀

Explore over 100,000 other datasets and pre-trained models on https://universe.roboflow.com. 🌍

🏷️ Annotation Format

Fractured bones in this dataset are annotated in YOLOv8 format, which is widely used for object detection tasks. 🎯

📚 Categories

Computer Vision, Medical Imaging, Deep Learning

🔗 Related Links

• Article citing this dataset: https://link.springer.com/article/10.1007/s10489-024-05588-7#citeas 📰

📜 Licence

This dataset is released under the Creative Commons Attribution 4.0 International License (CC BY 4.0). This means you are free to share and adapt the material for any purpose, even commercially, as long as you give appropriate credit, provide a link to the license, and indicate if changes were made. 📄

✍️ Citation

Parvin, Shahnaj (2024), “Human Bone Fractures Multi-modal Image Dataset (HBFMID)”, Mendeley Data, V1, doi: 10.17632/xwfs6xbk47.1

🏢 Institution

American International University Bangladesh 🇧🇩

Cauliflower Diseases: Black Rot (Original Images=103, Augmented Images=618) Card Rot (Original Images=105, Augmented Images=630) Tobbaco_Caterpilar (Original Images=100, Augmented Images=600) Yellow Virus (Original Images=100, Augmented Images=600)

Data Augmentation Techniques: A.HorizontalFlip(p=0.5), # 50% chance to flip horizontally A.VerticalFlip(p=0.3), # 30% chance to flip vertically A.RandomBrightnessContrast(p=0.3), # Adjust brightness & contrast randomly A.Rotate(limit=30, p=0.5), # Rotate between -30 to +30 degrees A.Affine(shear=(-20, 20), p=0.5), # Apply shear transformation in both X & Y directions A.GaussNoise(var_limit=(10.0, 50.0), p=0.3) # Add random noise

The SDFVD 2.0 is an augmented extension of the original SDFVD dataset, which originally contained 53 real and 53 fake videos. This new version has been created to enhance the diversity and robustness of the dataset by applying various augmentation techniques like horizontal flip, rotation, shear, brightness and contrast adjustment, additive gaussian noise, downscaling and upscaling to the original videos. These augmentations help simulate a wider range of conditions and variations, making the dataset more suitable for training and evaluating deep learning models for deepfake detection. This process has significantly expanded the dataset resulting in 461 real and 461 forged videos, providing a richer and more varied collection of video data for deepfake detection research and development. Dataset Structure The dataset is organized into two main directories: real and fake, each containing the original and augmented videos. Each augmented video file is named following the pattern: ‘

An updated version of the MSLD dataset, MSLD v2.0 has been released after being verified by an expert dermatologist!

For details, check our GitHub repo!

Context

The recent monkeypox outbreak has become a global healthcare concern owing to its rapid spread in more than 65 countries around the globe. To obstruct its expeditious pace, early diagnosis is a must. But the confirmatory Polymerase Chain Reaction (PCR) tests and other biochemical assays are not readily available in sufficient quantities. In this scenario, computer-aided monkeypox identification from skin lesion images can be a beneficial measure. Nevertheless, so far, such datasets are not available. Hence, the "Monkeypox Skin Lesion Dataset (MSLD)" is created by collecting and processing images from different means of web-scrapping i.e., from news portals, websites and publicly accessible case reports.

The creation of "Monkeypox Image Lesion Dataset" is primarily focused on distinguishing the monkeypox cases from the similar non-monkeypox cases. Therefore, along with the 'Monkeypox' class, we included skin lesion images of 'Chickenpox' and 'Measles' because of their resemblance to the monkeypox rash and pustules in initial state in another class named 'Others' to perform binary classification.

Content

There are 3 folders in the dataset.

1) Original Images: It contains a total number of 228 images, among which 102 belongs to the 'Monkeypox' class and the remaining 126 represents the 'Others' class i.e., non-monkeypox (chickenpox and measles) cases.

2) Augmented Images: To aid the classification task, several data augmentation methods such as rotation, translation, reflection, shear, hue, saturation, contrast and brightness jitter, noise, scaling etc. have been applied using MATLAB R2020a. Although this can be readily done using ImageGenerator/other image augmentors, to ensure reproducibility of the results, the augmented images are provided in this folder. Post-augmentation, the number of images increased by approximately 14-folds. The classes 'Monkeypox' and 'Others' have 1428 and 1764 images, respectively.

3) Fold1: One of the three-fold cross validation datasets. To avoid any sort of bias in training, three-fold cross validation was performed. The original images were split into training, validation and test set(s) with the approximate proportion of 70 : 10 : 20 while maintaining patient independence. According to the commonly perceived data preparation practice, only the training and validation images were augmented while the test set contained only the original images. Users have the option of using the folds directly or using the original data and employing other algorithms to augment it.

Additionally, a CSV file is provided that has 228 rows and two columns. The table contains the list of all the ImageID(s) with their corresponding label.

Web Application

Since monkeypox is demonstrating a very rapid community transmission pattern, a consumer-level software is truly necessary to increase awareness and encourage people to take rapid action. We have developed an easy-to-use web application named Monkey Pox Detector using the open-source python streamlit framework that uses our trained model to address this issue. It makes predictions on whether or not to see a specialist along with the prediction accuracy. Future updates will benefit from the user data we continue to collect and use to improve our model. The web app has a flask core, so that it can be deployed cross-platform in the future.

Learn more at our GitHub repo!

Citation

If this dataset helped your research, please cite the following articles:

Ali, S. N., Ahmed, M. T., Paul, J., Jahan, T., Sani, S. M. Sakeef, Noor, N., & Hasan, T. (2022). Monkeypox Skin Lesion Detection Using Deep Learning Models: A Preliminary Feasibility Study. arXiv preprint arXiv:2207.03342.

@article{Nafisa2022, title={Monkeypox Skin Lesion Detection Using Deep Learning Models: A Preliminary Feasibility Study}, author={Ali, Shams Nafisa and Ahmed, Md. Tazuddin and Paul, Joydip and Jahan, Tasnim and Sani, S. M. Sakeef and Noor, Nawshaba and Hasan, Taufiq}, journal={arXiv preprint arXiv:2207.03342}, year={2022} }

Ali, S. N., Ahmed, M. T., Jahan, T., Paul, J., Sani, S. M. Sakeef, Noor, N., Asma, A. N., & Hasan, T. (2023). A Web-based Mpox Skin Lesion Detection System Using State-of-the-art Deep Learning Models Considering Racial Diversity. arXiv preprint arXiv:2306.14169.

@article{Nafisa2023, title={A Web-base...

Background Information

This dataset was curated and annotated by Mohamed Attia.

The original dataset (v1) is composed of 451 images of various pills that are present on a large variety of surfaces and objects. https://i.imgur.com/shZh1DV.jpeg" alt="Example of an Annotated Image from the Dataset">

The dataset is available under the Public License.

Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

Dataset Versions

Version 1 (v1) - 451 images

• Preprocessing: Auto-Orient and Resize (Stretch to 416x416)
• Augmentations: No augmentations applied
• Training Metrics: This version of the dataset was not trained

Version 2 (v2) - 1,083 images

• Preprocessing: Auto-Orient, Resize (Stretch to 416x416), all classes remapped (Modify Classes) to "pill"
• Augmentations: 90° Rotate: Clockwise, Counter-Clockwise, Upside Down Crop: 0% Minimum Zoom, 77% Maximum Zoom Rotation: Between -45° and +45° Shear: ±15° Horizontal, ±15° Vertical Hue: Between -22° and +22° Saturation: Between -27% and +27% Brightness: Between -33% and +33% Exposure: Between -25% and +25% Blur: Up to 3px Noise: Up to 5% of pixels Cutout: 3 boxes with 10% size each Mosaic: Applied Bounding Box: Brightness: Between -25% and +25%
• Training Metrics: Trained from the COCO Checkpoint in Public Models ("transfer learning") on Roboflow
  • mAP = 91.4%, precision = 61.1%, recall = 93.9%

Version 3 (v3) - 1,083 images

• Preprocessing: Auto-Orient, Resize (Stretch to 416x416), all classes remapped (Modify Classes) to "pill"
• Augmentations: 90° Rotate: Clockwise, Counter-Clockwise, Upside Down Crop: 0% Minimum Zoom, 77% Maximum Zoom Rotation: Between -45° and +45° Shear: ±15° Horizontal, ±15° Vertical Hue: Between -22° and +22° Saturation: Between -27% and +27% Brightness: Between -33% and +33% Exposure: Between -25% and +25% Blur: Up to 3px Noise: Up to 5% of pixels Cutout: 3 boxes with 10% size each Mosaic: Applied Bounding Box: Brightness: Between -25% and +25%
• Training Metrics: Trained from "scratch" (no transfer learning employed) on Roboflow
  • mAP = 84.3%, precision = 53.2%, recall = 86.7%

Version 4 (v4) - 451 images

• Preprocessing: Auto-Orient, Resize (Stretch to 416x416), all classes remapped (Modify Classes) to "pill"
• Augmentations: No augmentations applied
• Training Metrics: This version of the dataset was not trained

Version 5 (v5) - 496 images

• Preprocessing: Auto-Orient, all classes remapped (Modify Classes) to "pill", Isolate Objects
  • The Isolate Objects preprocessing step was added to convert this object detection project into a suitable format for export in OpenAI's CLIP annotation format so that it could be used as a classifcation model (classification dataset available here: https://universe.roboflow.com/mohamed-attia-e2mor/pill-classification)

Mohamed Attia - LinkedIn

The original dataset is from https://www.kaggle.com/datasets/andyczhao/covidx-cxr2

The data is separated based on the .txt file (see link) into positive and negative.

Data Augmentation Code

from tensorflow.keras.preprocessing.image import ImageDataGenerator

from tensorflow.keras.preprocessing.image import ImageDataGenerator

datagen = ImageDataGenerator(
  rescale=1./255,        # Normalize
  rotation_range=20,      # Rotation reference
  zoom_range=0.2,        # Zoom reference
  width_shift_range=0.2,    # wrap
  height_shift_range=0.2,    # wrap
  shear_range=0.2,       # Add shear transformation
  brightness_range=(0.7, 1.3), # Wider brightness adjustment - reference 0.3
  horizontal_flip=True,
  fill_mode='nearest'
)

# Counts
current_count = len(os.listdir(input_dir))
target_count = 57199
required_augmented_count = target_count - current_count

print(f"Original negatives: {current_count}")
print(f"Required augmented images: {required_augmented_count}")

# augmenting ...
augmented_count = 0
max_augmentations_per_image = 10 #I used 5 and 10, this dataset was generated with 10

for img_file in os.listdir(input_dir):
  img_path = os.path.join(input_dir, img_file)
  img = load_img(img_path, target_size=(480, 480)) # 480 by 480 referring to reference.
  img_array = img_to_array(img)
  img_array = img_array.reshape((1,) + img_array.shape)

  # Generate multiple augmentations per image
  i = 0
  for batch in datagen.flow(
    img_array,
    batch_size=1,
    save_to_dir=output_dir,
    save_prefix='aug',
    save_format='jpeg'
  ):
    i += 1
    augmented_count += 1
    if i >= max_augmentations_per_image:
      break
    if augmented_count >= required_augmented_count:
      break

  if augmented_count >= required_augmented_count:
    break

I tried using different max_augmentations_per_image, or without setting this parameter; both ways generated augmented data (around 9,000) ...

positive_balanced: ```python random.seed(42)

Total negative samples

target_count = 20579

all_positive_images = os.listdir(positive_dir) selected_positive_images = random.sample(all_positive_images, target_count) ```

Background Information

This dataset was curated and annotated by Ilyes Talbi, Head of La revue IA, a French publication focused on stories of machine learning applications.

Main objetive is to identify if soccer (futbol) players, the referree and the soccer ball (futbol).

The original custom dataset (v1) is composed of 163 images. * Class 0 = players * Class 1 = referree * Class 2 = soccer ball (or futbol)

The dataset is available under the Public License.

Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

Dataset Versions

Version 7 (v7) - 163 images (raw images)

• Preprocessing: Auto-Orient, Modify Classes: 3 remapped, 0 dropped
  • Modified Classes: Class 0 = players, Class 1 = referree, Class 2 = futbol
• Augmentations: No augmentations applied
• Training Metrics: This version of the dataset was not trained

Version 2 (v2) - 163 images

• Preprocessing: Auto-Orient and Resize (Stretch to 416x416)
• Augmentations: No augmentations applied
• Training Metrics: This version of the dataset was not trained

Version 3 (v3) - 391 images

• Preprocessing: Auto-Orient and Resize (Stretch to 416x416), Modify Classes: 3 remapped, 0 dropped
  • Modified Classes: Class 0 = players, Class 1 = referree, Class 2 = futbol
• Augmentations:
  • Outputs per training example: 3
  • Rotation: Between -25° and +25°
  • Shear: ±15° Horizontal, ±15° Vertical
  • Brightness: Between -25% and +25%
  • Blur: Up to 0.75px
  • Noise: Up to 1% of pixels
  • Bounding Box: Blur: Up to 0.5px
• Training Metrics: 86.4%mAP, 51.8% precision, 90.4% recall

Version 4 (v4) - 391 images

• Preprocessing: Auto-Orient and Resize (Stretch to 416x416), Modify Classes: 3 remapped, 0 dropped
  • Modified Classes: Class 0 = players, Class 1 = referree, Class 2 = futbol
• Augmentations:
  • Outputs per training example: 3
  • Rotation: Between -25° and +25°
  • Shear: ±15° Horizontal, ±15° Vertical
  • Brightness: Between -25% and +25%
  • Blur: Up to 0.75px
  • Noise: Up to 1% of pixels
  • Bounding Box: Blur: Up to 0.5px
• Training Metrics: 84.6% mAP, 52.3% precision, 85.3% recall

Version 5 (v5) - 391 images

• Preprocessing: Auto-Orient and Resize (Stretch to 416x416), Modify Classes: 3 remapped, 2 dropped
  • Modified Classes: Class 0 = players, Class 1 = referree, Class 2 = futbol
    • Only Class 0, which was remapped to players was included in this version
• Augmentations:
  • Outputs per training example: 3
  • Rotation: Between -25° and +25°
  • Shear: ±15° Horizontal, ±15° Vertical
  • Brightness: Between -25% and +25%
  • Blur: Up to 0.75px
  • Noise: Up to 1% of pixels
  • Bounding Box: Blur: Up to 0.5px
• Training Metrics: Trained from the COCO Checkpoint in Public Models ("transfer learning") on Roboflow
  • 98.8%mAP, 76.3% precision, 99.2% recall

Version 6 (v6) - 391 images

• Preprocessing: Auto-Orient and Resize (Stretch to 416x416), Modify Classes: 3 remapped, 2 dropped
  • Modified Classes: Class 0 = players, Class 1 = referree, Class 2 = futbol
    • Only Class 0, which was remapped to players was included in this version
• Augmentations:
  • Outputs per training example: 3
  • Rotation: Between -25° and +25°
  • Shear: ±15° Horizontal, ±15° Vertical
  • Brightness: Between -25% and +25%
  • Blur: Up to 0.75px
  • Noise: Up to 1% of pixels
  • Bounding Box: Blur: Up to 0.5px
• Training Metrics: Trained from Scratch (no transfer learning employed)
  • 95.5%mAP, 67.8% precision, 95.5% recall

Ilyes Talbi - LinkedIn | La revue IA

On February 8, 2021, Deception Island Chinstrap penguin colonies were photographed during the PiMetAn Project XXXIV Spanish Antarctic campaign using unmanned aerial vehicles (UAV) at a height of 30m. From the obtained imagery, a training dataset for penguin detection from aerial perspective was generated. The penguin species is the Chinstrap penguin (Pygoscelis antarcticus). The dataset consists of three folders: "train", containing 531 images, intended for model training; "valid", containing 50 images, intended for model validation; and "test", containing 25 images, intended for model testing. In each of the three folders, an additional .csv file is located, containing labels (x,y positions and class names for every penguin in the images), annotated in Tensorflow Object Detection format. There is only one annotation class: Penguin. All 606 images are 224x224 px in size, and 96 dpi. The following augmentation was applied to create 3 versions of each source image: * Random shear of between -18° to +18° horizontally and -11° to +11° vertically This dataset was annotated and exported via www.roboflow.com The model Faster R-CNN64 with ResNet-101 backbone was used to perform object detection tasks. Training and evaluation tasks were performed using the TensorFlow 2.0 machine learning platform by Google.

This dataset presents an assortment of high-resolution images that exhibit six well-known banana varieties procured from two distinct regions in Bangladesh. These bananas were thoughtfully selected from rural orchards and local markets, providing a diverse and comprehensive representation. The dataset serves as a visual reference, offering a thorough portrayal of the distinct characteristics of these banana types, which aids in their precise classification. It encompasses six distinct categories, namely, Shagor, Shabri, Champa, Anaji, Deshi, and Bichi, with a total of 1166 original images and 6000 augmented JPG images. These images were diligently captured during the period from August 01 to August 15, 2023. The dataset includes two variations: one with raw images and the other with augmented images. Each variation is further categorized into six separate folders, each dedicated to a specific banana variety. The images are of non-uniform dimensions and have a resolution of 4608 × 3456 pixels. Due to the high resolution, the initial file size amounted to 4.08 GB. Subsequently, data augmentation techniques were applied, as machine vision deep learning models require a substantial number of images for effective training. Augmentation involves transformations like scaling, shifting, shearing, zooming, and random rotation. Specific augmentation parameters included rotations within a range of 1° to 40°, width and height shifts, zoom range, and shear ranges set at 0.2. As a result, an additional 1000 augmented images were generated from the original images in each category, resulting in a dataset comprising a total of 6000 augmented images (1000 per category) with a data size of 4.73 GB.

Background: The relationship between maxillary sinus pneumatization and respiratory-induced fluid mechanics remains unclear. The purpose of this study was to simulate and measure the respiratory-induced mechanical stimulation at the sinus floor under different respiratory conditions and to investigate its potential effect on the elevated sinus following sinus-lifting procedures.Methods: The nasal airway together with the bilateral maxillary sinuses of the selected patient was segmented and digitally modeled from a computed tomographic image. The sinus floors of the models were elevated by simulated sinus augmentations using computer-aided design. The numerical simulations of sinus fluid motion under different respiratory conditions were performed using a computational fluid dynamics (CFD) algorithm. Sinus wall shear stress and static pressure on the pre-surgical and altered sinus floors were examined and quantitatively compared.Results: Streamlines with minimum airflow velocity were visualized in the sinus. The sinus floor pressure and the wall shear stress increased with the elevated inlet flow rate, but the magnitude of these mechanical stimulations remained at a negligible level. The surgical technique and elevated height had no significant influence on the wall pressure and the fluid mechanics.Conclusion: This study shows that respiratory-induced mechanical stimulation in the sinus floor is negligible before and after sinus augmentation.

The UAVs-TEBDE (Turkey Earthquake Building Damage Estimation) dataset is a high-resolution aerial imagery collection developed to support AI-based post-earthquake damage assessment using deep learning and computer vision. Created in response to the 2023 Turkey earthquakes, the dataset provides annotated building imagery specifically curated for multi-class classification of structural integrity.

The original dataset consists of 1,636 images, each categorized into one of three damage levels:

Collapsed – Complete structural failure or irreparable destruction (474 images)

Damaged – Partial failure with visible deformation or cracking (664 images)

Intact – Structurally stable with no visible signs of damage (498 images)

Imagery was collected using a hybrid acquisition strategy combining:

UAV field missions conducted in the immediate aftermath of the 2023 Turkey earthquakes

Publicly available sources, including:

• Online media platforms (e.g., YouTube, news broadcasts) -Stock repositories (e.g., Shutterstock, Stock) -Open datasets (e.g., Kaggle, GitHub)

This multi-source approach ensures a diverse representation of building types, materials, damage patterns, and environmental conditions (e.g., variations in lighting, resolution, and viewing angles), enhancing the dataset’s generalizability for real-world disaster response scenarios.

Data Augmentation Strategy To address the limited sample size and improve model robustness, a comprehensive image augmentation pipeline was applied to the original dataset. This process generated synthetic but realistic image variants while preserving core structural features.

The augmentation parameters used include:

*Rotation Range: ±160° *Width Shift: 0.2 *Height Shift: 0.2 *Shear Range: 0.2 *Zoom Range: 0.25 *Horizontal Flip: Enabled *Fill Mode: Reflect *Constant Fill Value: 125 *Batch Size: 32 *Augmentation Cycles: 200+

This augmentation strategy increased the total number of samples to 5,500 images per class, resulting in a final dataset size of 16,500 images:

Collapsed: 5,500 images

Damaged: 5,500 images

Intact: 5,500 images

This enhanced version of UAVs-TEBDE offers a balanced, diverse, and high-quality benchmark for training and evaluating advanced building damage detection models.

Code Availability The related model architecture and training pipeline, including the SCA_HMDA attention module, Vision Transformer, and data augmentation routines, are openly available in the following GitHub repository: https://github.com/najmulmowla1/Earthquake-Building-Damage-Detection

The dataset contains soybean crop adult insect augmented cropped Images ( a total of 7306 images) in JPG format. There were considered four kind of soybean crop insect’s images as:

1. A folder Augmented_Cropped_Eocanthecona_Bug_A contains 1802 images of Eocanthecona Bug. The augmentation methods and corresponding images are as:

Brightness (contrast):201, flipping (horizontal):300, rotation (45 degree): 301, saturation: 102, scaling:300, shearing:299, translation:299

1. A folder Augmented_Cropped_Tobacco_Caterpillar_B contains 1812 images of Tobacco Caterpillar. The augmentation methods and corresponding images are as:

Brightness (contrast):200, flipping (horizontal):300, rotation (45 degree): 313, saturation: 101, scaling:300, shearing:299, translation:299

1. A folder Augmented_Cropped_Red_Hairy_Catterpillar_C Contains 1798 images of Red Hairy Caterpillar. The augmentation methods and corresponding images are as:

Brightness (contrast):200, flipping (horizontal):300, rotation (45 degree): 300, saturation: 100, scaling:300, shearing:299, translation:299

1. A folder Augmented_Cropped_Larva_Spodoptera_D contains 1894 images of Larva Spodoptera. The augmentation methods and corresponding images are as:

Brightness (contrast):201, flipping (horizontal):300, rotation (45 degree): 300, saturation: 101, scaling:300, shearing:299, translation:393

1. Initially, all the images were taken from the following two datasets and then further applied various augmentation methods (brightness, flipping, rotation, saturation, scaling, shearing, translation) to enhance the dataset size:

(i). Tiwari, Vivek; Saxena, Ravi R; Ojha, Muneendra (2020): Soybean Crop Insect Raw Image Dataset_V1 with Bounding boxes for Classification and Localization. figshare. Dataset. https://doi.org/10.6084/m9.figshare.13077221.v3

(ii) Tiwari, Vivek; Saxena, Ravi R; Ojha, Muneendra (2020): Soybean Crop Insect Processed (Cropped) Image Dataset_V1 for Classification. figshare. Dataset. https://doi.org/10.6084/m9.figshare.13078883

The dataset was developed by the authors under collaborative work between Indira Gandhi Krishi Vishwavidyalaya (IGKV), Raipur (CG), and DSPM IIIT Naya Raipur (CG), India

IntroductionLow bone density and lack of medial support are the two most important factors affecting the stability of locking plate fixation for osteoporotic proximal humeral fractures (PHFs). This study aimed to compare the biomechanical characteristics of PHILOS locking plates combined with calcar screws, bone cement, fibular allografts, and medial locking plate support strategies for treating osteoporotic PHFs with medial column instability.MethodsA three-part osteoporotic PHF (AO 11-B3.2) model with metaphyseal loss was generated using 40 synthetic humeri and fixed via four distinct medial support strategies. All models were mechanically tested to quantify the mechanical characteristics. Subsequently, finite element models were created for each biomechanical test case. The stress distribution and displacement of the four different fixation structures were analyzed using finite element analysis.ResultsThe results demonstrated that the PHILOS locking plate combined with the medial locking plate, exhibited the greatest stability when subjected to axial, shear, and torsional loading. Furthermore, the PHILOS locking plate combined with bone cement showed structural stability similar to that of the PHILOS locking plate combined with fibular allograft but with lower stress levels on the fracture surface.DiscussionIn conclusion, the PLP-MLP fixation structure showed superior biomechanical properties under axial, shear, and torsional loading compared to other medial support methods. Repairing the medial support when treating osteoporotic PHFs with medial column instability can enhance the mechanical stability of the fracture end in both the short and long term.

This dataset includes dental OPG X-rays collected from three different dental clinics. This dataset can be used for tasks like object detection, image analysis, disease classification, and segmentation. It has two folders: one with 232 original images and their labels in JSON format, and another with 604 augmented images and labels. The augmentations were done using Roboflow and involved techniques like flipping, rotation, resizing, shear, mosaic augmentation, and bounding box exposure. The augmented data is split into training, validation, and testing sets in an 80:10:10 ratio.

Dataset collection: • Source: Prescription Point Ltd, Lab Aid Specialized Hospital, Ibn Sina Diagnostic and Imaging Center. • Capture Method: Using android phone camera. • Anonymization: All data were rigorously anonymized to maintain confidentiality and privacy. • Informed Consent: All patients provided their consent in accordance with the dental ethical principles.

Dataset composition: • Total Participants: 232 Male and female patients aged 10 years or older.

Variables: • Healthy Teeth: 223 • Caries: 119 • Impacted Teeth: 87 • Broken Down Crow/ Root: 52 • Infection: 23 • Fractured Teeth: 13

Context

I made this data annotation for conference paper . I try to make an application that will be fast and light enough to deploy in any cutting edge device while maintaining a good accuracy like any state-of-the-art model.

Data Details

The following pre-processing was applied to each image: * Auto-orientation of pixel data (with EXIF-orientation stripping) * Resize to 416x416 (Stretch)

The following augmentation was applied to create 3 versions of each source image in trainig set images: * 50% probability of horizontal flip * 50% probability of vertical flip * Equal probability of one of the following 90-degree rotations: none, clockwise, counter-clockwise, upside-down * Randomly crop between 0 and 7 percent of the image * Random rotation of between -40 and +40 degrees * Random shear of between -29° to +29° horizontally and -15° to +15° vertically * Random exposure adjustment of between -34 and +34 percent * Random Gaussian blur of between 0 and 1.5 pixels * Salt and pepper noise was applied to 4 percent of pixels

Acknowledgements

A big shoutout to Massey University for making this dataset public. The original dataset Link is : here , Please keep in mind that the original dataset maybe updated from time to time. However, I don't intend to update this annotated version.

The Jujube Diseases Identification Image Dataset provides high-resolution images and precise measurements of various jujube diseases, capturing key attributes such as size, color, and texture. This dataset serves as a valuable resource for agricultural research, offering detailed morphological and spectral data to aid in disease detection, classification, and management. Researchers can leverage these insights to refine cultivation techniques and enhance resistance to pests and diseases. Dataset Overview • Total Images: 1000 original images, captured in real-world field conditions • Augmented Images: 6000 (generated using data augmentation techniques) Augmentation Pipeline To enhance dataset diversity and improve model generalization, the following augmentation techniques were applied: 1. Horizontal flipping (50% probability) 2. Vertical flipping (30% probability) 3. Random brightness and contrast adjustments 4. Rotation within a range of -30° to +30° 5. Shear transformation along both X and Y axes 6. Addition of random noise This enriched dataset provides a robust foundation for developing deep learning models for automated jujube disease identification, contributing to precision agriculture and improved disease management strategies.

Context

During the initial peak outbreak phase of Mpox, a significant challenge emerged due to the absence of a publicly available reliable dataset for the detection of Mpox. The rapid escalation of Mpox cases, with its potential spread reaching Europe and America as highlighted by the World Health Organization, along with emerging possibilities of Mpox cases in Asian countries, underscored the urgency of implementing computer-assisted detection as a critical tool. In this context, the immediate diagnosis of Mpox became an increasingly challenging endeavor. As the possibility of a Mpox outbreak loomed over densely populated countries like Bangladesh, the limitations of our available resources rendered rapid diagnosis unattainable. Hence, the dire need for computer-assisted detection methods became apparent.

To address this pressing need, the development of computer-assisted methods demanded an ample amount of diverse data, including skin lesion images of Mpox from individuals of different sexes, ethnicities, and skin tones. However, the scarcity of available data posed a considerable obstacle in this endeavor. In response to this critical situation, our research group took the initiative to develop one of the earliest datasets (MSLD) specifically tailored for Mpox, encompassing various classes including non-Mpox samples.

From June 2022 to May 2023, the Mpox Skin Lesion Dataset (MSLD) has undergone two iterations, resulting in the current version, MSLD v2.0. The previous version included two classes: "Mpox" and "Others" (non-Mpox), with the "Others" class comprising skin lesion images of chickenpox and measles, chosen for their similarity to Mpox. Building upon the limitations identified in the initial release, we have developed an enhanced and more comprehensive version, MSLD v2.0. This updated dataset encompasses a wider range of classes and provides a more diverse set of images suitable for multi-class classification.

MSLD v2.0 comprises images from six distinct classes, namely Mpox (284 images), Chickenpox (75 images), Measles (55 images), Cowpox (66 images), Hand-foot-mouth disease or HFMD (161 images), and Healthy (114 images). The dataset includes 755 original skin lesion images sourced from 541 distinct patients, ensuring a representative sample. Importantly, the latest version has received endorsement from professional dermatologists and obtained approval from appropriate regulatory authorities.

Content

The dataset is organized into two folders:

Original Images: This folder includes a subfolder named "FOLDS" containing five folds (fold1-fold5) for 5-fold cross-validation with the original images. Each fold has separate folders for the test, train, and validation sets.

Augmented Images: To enhance the classification task, various data augmentation techniques, such as rotation, translation, reflection, shear, hue, saturation, contrast, brightness jitter, noise, and scaling, were applied using MATLAB R2020a. To ensure result reproducibility, the augmented images are provided in this folder. It contains a subfolder called "FOLDS_AUG" with augmented images of the train sets from each fold in the "FOLDS" subfolder of the "Original Images". The augmentation process resulted in an approximate 14-fold increase in the number of images.

Naming Convention of the Images

Each image is assigned a name following the format of DiseaseCode_PatientNumber_ImageNumber. The corresponding disease codes assigned to each of the six disease classes are - Mpox -> MKP, Chickenpox -> CHP, Cowpox -> CWP, Measles -> MSL, Hand,foot and mouth disease -> HFMD, Healthy -> HEALTHY. Assignment of the keywords is illustrated in the provided image "Keywords.jpg". For instance, an image named "MKP_17_01" indicates that it belongs to the Mpox class and is the first image captured from a patient with the ID 17.

Data organization

The dataset includes an Excel file named "**datalog.xlsx**" consisting of 5 sheets (Sheet1-5), with each sheet corresponding to a specific fold (fold1-5). Each sheet contains three columns: train, validation, and test. These columns contain the names of the images belonging to the respective train, validation, and test sets for a particular fold.

Web Application

Since we intend to build an end to end solution - starting with dataset creation and ending with a live web app, a prototype of the web-app has already been developed using the open-source python streamlit framework with a flask core and has been hosted in the streamlit provided server for better user experience. In the app, Skin Lesion Detector, users can get, not only a suggestion but also the accuracy of the suggestion.

The codes required to build and train the model, all ...

This Bangladeshi Brain Cancer MRI Dataset is a large dataset of Magnetic Resonance Imaging (MRI) images created to aid researchers in medical diagnosis, especially for brain cancer research. This collection contains a total of 1600 raw photos (every class have 400 raw images) after augmentation it contains total 6000 images, which are wisely divided into four main categories as:

Glioma -1500 images

Meningioma -1500 images

Pituitary-1500 images

No Tumor-1500 images

All the images in this dataset were collected from different hospitals around Bangladesh. It brought diversity and representation into the sample. To make the images compatible with various image processing, machine learning and deep-learning pipelines as possible they are then resized to a standardize size of 512×512.

This dataset is incredibly significant since high-quality data, such as medical imaging data, are few and difficult to obtain, particularly in the context of brain cancer. Assume that four prominent doctors collaborate on data collection in order to give more accurate and helpful content. It made it feasible. The cooperation emphasizes the dataset's potential to improve medical practice today by providing a dependable supply of diagnoses for use in diagnostic tool creation and testing within current medicine.

This dataset can be used by researchers and practitioners for a variety of applications such as Dense net 201, yolov8x/s, CNN, resnet50v2, VGG-16, MobilenetV2 etc.

Image Processing Details:

Images are randomly rotated within a range of 45 degrees. (rotation range=45)

Images are horizontally shifted by up to 20% of the width of the image. (width_shift_range=0.2)

Images are vertically shifted by up to 20% of the height of the image. (height_shift_range=0.2)

Shear transformation is applied to the image within a range of 20%. (shear range=0.2)

Images are randomly zoomed in or out by up to 20%. (zoom range=0.2)

Images are randomly flipped horizontally. (horizontal flip=True)

When transformations like rotations or shifts leave empty areas in the image, they are filled in by the nearest pixel values. (fill mode='nearest')

Hospital List(for Data Collection):

Ibn Sina Medical College, Kollanpur, 1, 1-B Mirpur Rd, Dhaka 1207

Dhaka Medical College & Hospital, Secretariat Rd, Dhaka 1000

Cumilla Medical College, Kuchaitoli, Dr. Akhtar Hameed Khan Road, Cumilla 3500, Bangladesh

Supervisor & investigator:

Md. Mizanur Rahman

Lecturer,

Computer Science and Engineering

Daffodil International University

Dhaka, Bangladesh

mizanurrahman.cse@diu.edu.bd

Data Collectors:

Md Shahriar Mannan Prottoy

Mahtab Chowdhury

Redwan Rahman

Azim Ullah Tamim