Data Augmented Flip Exposure

## Overview

Data Augmented Flip Exposure is a dataset for object detection tasks - it contains Damage annotations for 2,397 images.

## 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.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

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 ])

Data associated with research towards a surrogate machine learning model for the Advanced Gas-cooled Reactor (AGR). This data was generated using the Parmec software package [1] and can be used to train machine learning models using the Surrogate Machine Optimisation and Learning (SMOL) framework [2]. Visit the aforementioned repository, clone the code, then download the files into repository folder.

If you are not planning on working with data augmentation, exclude the files with flip and rotate in the title, e.g. dataset_flip_13_rotate_123_cases.pkl.

1. Koziara, T., 2019. Parmec documentation. URL: https://parmes.org/parmec/index.html [Online; accessed 05-August-2022].

2. Github Repository. URL: https://gitlab.cs.man.ac.uk/q59494hj/parmec_agr_ml_surrogate [Online; accessed 05-August-2022].

This dataset is a meticulously curated dataset designed for infant facial emotion recognition, featuring four primary emotional expressions: Angry, Cry, Laugh, and Normal. The dataset aims to facilitate research in machine learning, deep learning, affective computing, and human-computer interaction by providing a large collection of labeled infant facial images.

Primary Data (1600 Images): - Angry: 400 - Cry: 400 - Laugh: 400 - Normal: 400

Data Augmentation & Expanded Dataset (26,143 Images): To enhance the dataset's robustness and expand the dataset, 20 augmentation techniques (including HorizontalFlip, VerticalFlip, Rotate, ShiftScaleRotate, BrightnessContrast, GaussNoise, GaussianBlur, Sharpen, HueSaturationValue, CLAHE, GridDistortion, ElasticTransform, GammaCorrection, MotionBlur, ColorJitter, Emboss, Equalize, Posterize, FogEffect, and RainEffect) were applied randomly. This resulted in a significantly larger dataset with:

• Angry: 5,781
• Cry: 6,930
• Laugh: 6,870
• Normal: 6,562

Data Collection & Ethical Considerations: The dataset was collected under strict ethical guidelines to ensure compliance with privacy and data protection laws. Key ethical considerations include: 1. Ethical Approval: The study was reviewed and approved by the Institutional Review Board (IRB) of Daffodil International University under Reference No: REC-FSIT-2024-11-10. 2. Informed Parental Consent: Written consent was obtained from parents before capturing and utilizing infant facial images for research purposes. 3. Privacy Protection: No personally identifiable information (PII) is included in the dataset, and images are strictly used for research in AI-driven emotion recognition.

Data Collection Locations & Geographical Diversity: To ensure diversity in infant facial expressions, data collection was conducted across multiple locations in Bangladesh, covering healthcare centers and educational institutions:

1. 250-bed District Sadar Hospital, Sherpur (Latitude: 25.019405 & Longitude: 90.013733)
2. Upazila Health Complex, Baraigram, Natore (Latitude: 24.3083 & Longitude: 89.1700)
3. Char Bhabna Community Clinic, Sherpur (Latitude: 25.0188 & Longitude: 90.0175)
4. Jamiatul Amin Mohammad Al-Islamia Cadet Madrasa, Khagan, Dhaka (Latitude: 23.872856 & Longitude: 90.318947)

Face Detection Methodology: To extract the facial regions efficiently, RetinaNet—a deep learning-based object detection model—was employed. The use of RetinaNet ensures precise facial cropping while minimizing background noise and occlusions.

Potential Applications: 1. Affective Computing: Understanding infant emotions for smart healthcare and early childhood development. 2. Computer Vision: Training deep learning models for automated infant facial expression recognition. 3. Pediatric & Mental Health Research: Assisting in early autism screening and emotion-aware AI for child psychology. 4. Human-Computer Interaction (HCI): Designing AI-powered assistive technologies for infants.

Car-Highway Data Annotation Project

Introduction

In this project, we aim to annotate car images captured on highways. The annotated data will be used to train machine learning models for various computer vision tasks, such as object detection and classification.

Project Goals

• Collect a diverse dataset of car images from highway scenes.
• Annotate the dataset to identify and label cars within each image.
• Organize and format the annotated data for machine learning model training.

Tools and Technologies

For this project, we will be using Roboflow, a powerful platform for data annotation and preprocessing. Roboflow simplifies the annotation process and provides tools for data augmentation and transformation.

Annotation Process

1. Upload the raw car images to the Roboflow platform.
2. Use the annotation tools in Roboflow to draw bounding boxes around each car in the images.
3. Label each bounding box with the corresponding class (e.g., car).
4. Review and validate the annotations for accuracy.

Data Augmentation

Roboflow offers data augmentation capabilities, such as rotation, flipping, and resizing. These augmentations can help improve the model's robustness.

Data Export

Once the data is annotated and augmented, Roboflow allows us to export the dataset in various formats suitable for training machine learning models, such as YOLO, COCO, or TensorFlow Record.

Milestones

1. Data Collection and Preprocessing
2. Annotation of Car Images
3. Data Augmentation
4. Data Export
5. Model Training

Conclusion

By completing this project, we will have a well-annotated dataset ready for training machine learning models. This dataset can be used for a wide range of applications in computer vision, including car detection and tracking on highways.

This dataset provides an augmented version of the training images and corresponding segmentation masks from the PlantSeg v3 dataset (available at /kaggle/input/plantseg/plantsegv3/). The goal of this augmentation is to artificially expand the dataset size and diversity, aiming to help train more robust and generalizable machine learning models for plant segmentation tasks.

Dataset Contents & Generation:

• Source: Original training images (images/train) and annotations (annotations/train) from PlantSeg v3.
• Format: Paired images (.jpg or .png) and single-channel grayscale segmentation masks (saved as .png).
• Augmentation Strategy:
  • For each original image-mask pair successfully processed, 4 unique augmented versions were generated using the albumentations library.
  • The original, unmodified image-mask pairs are also included in the final dataset.
  • This results in approximately a 5x increase in the number of training examples (1 original + 4 augmented per source pair).
• Process: A Python script was used to apply a defined augmentation pipeline, save the results, and package them. Only image-mask pairs where both files were found and valid were processed.

File Structure:

The dataset archive (train_augmented.zip) contains two main directories:

• images/: Contains all training images (original and augmented). Original files have an _orig suffix, augmented files have _aug_N (where N is 0-3).
• masks/: Contains the corresponding segmentation masks, named to match their respective images.

Augmentation Pipeline:

A comprehensive pipeline from the albumentations library was applied. It combines various geometric (spatial) and pixel-level (color, noise, blur) transformations. Geometric transformations were applied consistently to both images and masks to maintain spatial alignment. Key transformation types include:

• Geometric: Flips, Rotations (90 degrees), Shift-Scale-Rotate, Elastic Transformation, Grid Distortion.
• Color & Pixel (Image Only): Brightness/Contrast, Hue/Saturation/Value adjustments, Gaussian Noise, various Blurs (Gaussian, Motion, Median), CLAHE.
• Other: Coarse Dropout (Cutout) applied to both image and mask.

See the detailed table below for specific transformations, parameters, and application probabilities.

Potential Use Cases:

This dataset is suitable for training and evaluating deep learning models for semantic or instance segmentation of plants, particularly leaves, under varied conditions simulated by the augmentations.

Acknowledgements:

This augmented dataset is derived from the PlantSeg v3 dataset. Please acknowledge the original dataset creators if you use this data.

Detailed Augmentation Pipeline

The following table details the transformations included in the albumentations.Compose pipeline used for generation. The overall Compose pipeline has p=1.0, but individual transformations are applied based on their respective probabilities (p).

Please note that the file msl-labeled-data-set-v2.1.zip below contains the latest images and labels associated with this data set.

Data Set Description

The data set consists of 6,820 images that were collected by the Mars Science Laboratory (MSL) Curiosity Rover by three instruments: (1) the Mast Camera (Mastcam) Left Eye; (2) the Mast Camera Right Eye; (3) the Mars Hand Lens Imager (MAHLI). With the help from Dr. Raymond Francis, a member of the MSL operations team, we identified 19 classes with science and engineering interests (see the "Classes" section for more information), and each image is assigned with 1 class label. We split the data set into training, validation, and test sets in order to train and evaluate machine learning algorithms. The training set contains 5,920 images (including augmented images; see the "Image Augmentation" section for more information); the validation set contains 300 images; the test set contains 600 images. The training set images were randomly sampled from sol (Martian day) range 1 - 948; validation set images were randomly sampled from sol range 949 - 1920; test set images were randomly sampled from sol range 1921 - 2224. All images are resized to 227 x 227 pixels without preserving the original height/width aspect ratio.

Directory Contents

images - contains all 6,820 images

class_map.csv - string-integer class mappings

train-set-v2.1.txt - label file for the training set

val-set-v2.1.txt - label file for the validation set

test-set-v2.1.txt - label file for the test set

The label files are formatted as below:

"Image-file-name class_in_integer_representation"

Labeling Process

Each image was labeled with help from three different volunteers (see Contributor list). The final labels are determined using the following processes:

If all three labels agree with each other, then use the label as the final label.

If the three labels do not agree with each other, then we manually review the labels and decide the final label.

We also performed error analysis to correct labels as a post-processing step in order to remove noisy/incorrect labels in the data set.

Classes

There are 19 classes identified in this data set. In order to simplify our training and evaluation algorithms, we mapped the class names from string to integer representations. The names of classes, string-integer mappings, distributions are shown below:

Class name, counts (training set), counts (validation set), counts (test set), integer representation

Arm cover, 10, 1, 4, 0

Other rover part, 190, 11, 10, 1

Artifact, 680, 62, 132, 2

Nearby surface, 1554, 74, 187, 3

Close-up rock, 1422, 50, 84, 4

DRT, 8, 4, 6, 5

DRT spot, 214, 1, 7, 6

Distant landscape, 342, 14, 34, 7

Drill hole, 252, 5, 12, 8

Night sky, 40, 3, 4, 9

Float, 190, 5, 1, 10

Layers, 182, 21, 17, 11

Light-toned veins, 42, 4, 27, 12

Mastcam cal target, 122, 12, 29, 13

Sand, 228, 19, 16, 14

Sun, 182, 5, 19, 15

Wheel, 212, 5, 5, 16

Wheel joint, 62, 1, 5, 17

Wheel tracks, 26, 3, 1, 18

Image Augmentation

Only the training set contains augmented images. 3,920 of the 5,920 images in the training set are augmented versions of the remaining 2000 original training images. Images taken by different instruments were augmented differently. As shown below, we employed 5 different methods to augment images. Images taken by the Mastcam left and right eye cameras were augmented using a horizontal flipping method, and images taken by the MAHLI camera were augmented using all 5 methods. Note that one can filter based on the file names listed in the train-set.txt file to obtain a set of non-augmented images.

90 degrees clockwise rotation (file name ends with -r90.jpg)

180 degrees clockwise rotation (file name ends with -r180.jpg)

270 degrees clockwise rotation (file name ends with -r270.jpg)

Horizontal flip (file name ends with -fh.jpg)

Vertical flip (file name ends with -fv.jpg)

Acknowledgment

The authors would like to thank the volunteers (as in the Contributor list) who provided annotations for this data set. We would also like to thank the PDS Imaging Note for the continuous support of this work.

This data set contains a total of 73,031 landmarks. 10,433 landmarks were detected and extracted from 180 HiRISE browse images, and 62,598 landmarks were augmented from 10,433 original landmarks. For each original landmark, we cropped a square bounding box that includes the full extent of the landmark plus a 30-pixel margin to left, right, top and bottom. Each cropped landmark was resized to 227x227 pixels, and then was augmented to generate 6 additional landmarks using the following methods: 90 degrees clockwise rotation 180 degrees clockwise rotation 270 degrees clockwise rotation Horizontal flip Vertical flip Random brightness adjustment

Description:

👉 Download the dataset here

This dataset consists of a diverse collection of images, tailored specifically for the task of Animal Image Classification Dataset in the domain of animal species. It contains 15 distinct folders, each corresponding to a unique animal class, with each folder representing the name of the animal species. The dataset is composed of a variety of images that have been preprocessed and prepared for use in machine learning applications.

Dataset Details:

Image Size: Each image in the dataset has been resized to dimensions of 224x224 pixels with 3 color channels (RGB), making them suitable for immediate use in neural networks.

Data Source: Images were sourced from publicly available databases on the web. They encompass various environments, lighting conditions, and angles, ensuring a rich and diverse representation of each animal class.

Classes: The dataset includes 15 animal classes such as cats, dogs, birds, elephants, lions, and more, with each class represented by images stored in its respective folder.

Download Dataset

Preprocessing and Augmentation:

The dataset underwent extensive preprocessing using OpenCV libraries, ensuring that all images were standardized to the same size. In addition to resizing, multiple augmentation techniques were applied to diversify the dataset and improve model generalization. These augmentations include:

Rotation: Random rotations applied to simulate different perspectives.

Flipping: Horizontal flips to account for variations in animal orientation.

Cropping: Random cropping to focus on various parts of the animal subjects.

Scaling: Minor scaling adjustments to simulate different zoom levels.

All preprocessing and augmentation were carried out to enhance the robustness of any model trained on this data, without the need for further augmentation steps. Therefore, the dataset is ready for immediate use in training deep learning models such as CNNs (Convolutional Neural Networks) or transfer learning models.

Applications:

This dataset is ideal for:

Image Classification: Train models to accurately classify different animal species.

Transfer Learning: Utilize pre-trained models to fine-tune performance on this dataset.

Computer Vision Research: Explore various computer vision tasks, such as animal identification, object detection, and species recognition.

Wildlife and Conservation Studies: Use the dataset to build Al systems capable of identifying animals in the wild for tracking and conservation efforts.

Potential Use Cases:

Education: For students and researchers to learn and experiment with animal classification using computer vision techniques.

Al and Machine Learning Competitions: A challenging dataset for machine learning competitions centered around image classification.

Mobile Applications: Can be used to develop apps for real-time animal identification using image recognition technology.

Dataset Format:

The dataset is structured for ease of use, with each folder containing images pertaining to a specific class. The file format is as follows:

Folder Structure: dataset/{class_name}/{image_files.jpg}

Image Type: JPEG/PNG

Annotations: No specific annotations are included, but each folder name serves as the label for the images within it.

This dataset is sourced from Kaggle.

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: ‘

The tree_detection_dataset is a subset of the original tree classification dataset, accessible at this link. This dataset includes only the original images, each with a resolution of 1200x1600. A further refined subset, yolo11, includes bounding box annotations specifically for object detection. All augmentations and processing were performed using Roboflow.

Dataset Configuration

The dataset comprises 2,716 images, divided as follows: - Training Set: 87% (2,376 images) - Validation Set: 8% (227 images) - Test Set: 5% (113 images)

To ensure uniformity in aspect ratio, all images have been resized to 640x640 pixels using a "fit" resizing approach, which may introduce black edges. Additionally, auto-orientation has been applied for consistency.

Data Preprocessing and Augmentation

Several augmentation techniques have been applied to enhance model robustness and generalization. Each training example generates three augmented outputs, expanding the diversity of the dataset. The augmentations include:

• Flip: Horizontal flipping.
• Rotation:
  • 90° rotations: Clockwise, counterclockwise, and upside-down.
  • Minor random rotations: Between -3° and +3° for slight variations.
• Grayscale: Applied to 12% of images to simulate different lighting conditions.
• Blur: Up to 1.5 pixels to mimic focus variations.
• Noise: Added to up to 0.1% of pixels to introduce subtle distortions.

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.

This extensive dataset presents a meticulously curated collection of low-resolution images showcasing 20 well-established rice varieties native to diverse regions of Bangladesh. The rice samples were carefully gathered from both rural areas and local marketplaces, ensuring a comprehensive and varied representation. Serving as a visual compendium, the dataset provides a thorough exploration of the distinct characteristics of these rice varieties, facilitating precise classification.

Dataset Composition

The dataset encompasses 18 distinct classes, encompassing Subol Lota, Bashmoti (Deshi), Ganjiya, Shampakatari, Sugandhi Katarivog, BR-28, BR-29, Paijam, Bashful, Lal Aush, BR-Jirashail, Gutisharna, Birui, Najirshail, Pahari Birui, Polao (Katari), Polao (Chinigura), Amon, Shorna-5, and Lal Binni. In total, the dataset comprises 4,730 original JPG images and 23,650 augmented images.

Image Capture and Dataset Organization

These images were captured using an iPhone 11 camera with a 5x zoom feature. Each image capturing these rice varieties was diligently taken between October 18 and November 29, 2023. To facilitate efficient data management and organization, the dataset is structured into two variants: Original images and Augmented images. Each variant is systematically categorized into 20 distinct sub-directories, each corresponding to a specific rice variety.

Original Image Dataset

The primary image set comprises 4,730 JPG images, uniformly sized at 853 × 853 pixels. Due to the initial low resolution, the file size was notably 268 MB. Employing compression through a zip program significantly optimized the dataset, resulting in a final size of 254 MB.

Augmented Image Dataset

To address the substantial image volume requirements of deep learning models for machine vision, data augmentation techniques were implemented. Transformations such as rotation (90° left, 90° right, 180° left) and flip were applied, generating an additional set of augmented images in every class, totaling 23,650 augmented images. These augmented images, also in JPG format and uniformly sized at 512 × 512 pixels, initially amounted to 781 MB. However, post-compression, the dataset was further streamlined to 699 MB.

Dataset Storage and Access

The raw and augmented datasets are stored in two distinct zip files, namely 'Original.zip' and 'Augmented.zip'. Both zip files contain 20 sub-folders representing a unique rice variety, namely 1_Subol_Lota, 2_Bashmoti, 3_Ganjiya, 4_Shampakatari, 5_Katarivog, 6_BR28, 7_BR29, 8_Paijam, 9_Bashful, 10_Lal_Aush, 11_Jirashail, 12_Gutisharna, 13_Red_Cargo,14_Najirshail, 15_Katari_Polao, 16_Lal_Biroi, 17_Chinigura_Polao, 18_Amon, 19_Shorna5, 20_Lal_Binni.

Please note that the file hirise-map-proj-v3_2.zip below contains the latest images and labels associated with this data set.

This dataset contains a total of 64,947 landmark images that were detected and extracted from HiRISE browse images, spanning 232 separate source images.

This set was formed from 10,815 original landmarks. Each original landmark was cropped to a square bounding box that included the full extent of the landmark plus a 30-pixel margin to the left, right, top, and bottom. Each landmark was then resized to 227x227 pixels. 9,022 of these images were then augmented to generate 6 additional landmarks using the following methods:

1. 90 degrees clockwise rotation
2. 180 degrees clockwise rotation
3. 270 degrees clockwise rotation
4. Horizontal flip
5. Vertical flip
6. Random brightness adjustment

The remaining 1,793 images were not augmented. Combining these with the 7*9,022 images, gives a total of 64,947 separate images.

Contents:
- map-proj-v3_2/: Directory containing individual cropped landmark images
- labels-map-proj-v3_2.txt: Class labels (ids) for each landmark image. File includes two columns separated by a space: filename, class_id

- labels-map-proj-v3_2_train_val_test.txt: Includes train/test/val labels and upsampling used for trained model. File includes three columns separated by a space: filename, class_id, set
- landmarks_map-proj-v3_2_classmap.csv: Dictionary that maps class ids to semantic names

Class Discussion:

We give a discussion of the various landmarks that make up our classes.

Bright dune and dark dune are two sand dune classes found on Mars. Dark dunes are completely defrosted, whereas bright dunes are not. Bright dunes are generally bright due to overlying frost and can exhibit black spots where parts of the dune are defrosting.

The crater class consists of crater images in which the diameter of the crater is greater than or equal to 1/5 the width of the image and the circular rim is visible for at least half the crater's circumference.

The slope streak class consists of images of dark flow-like features on slopes. These features are believed to be formed by a dry process in which overlying (bright) dust slides down a slope and reveals a darker sub-surface.

Impact ejecta refers to material that is blasted out from the impact of a meteorite or the eruption of a volcano. We also include cases in which the impact cleared away overlying dust, exposing the underlying surface. In some cases, the associated crater may be too small to see. Impact ejecta can also include lava that spilled out from the impact (blobby ("lobate") instead of blast-like), more like an eruption (triggered by the impact). Impact ejecta can be isolated, or they can form in clusters when the impactor breaks up into multiple fragments.

Spiders and Swiss cheese are phenomena that occur in the south polar region of Mars. Spiders have a central pit with radial troughs, and they are believed to form as a result of sublimation of carbon dioxide ice. This process can produce mineral deposits on top, which look like dark or light dust that highlights cracks in the CO2 ice. Spiders can resemble impact ejecta due to their radial troughs, but impact ejecta tends to have straight radial jets that fade as they get farther from the center. The spider class also includes fan-like features that form when a geyser erupts through the CO2 layer and the material is blown by the wind away from the cracks. Fans are typically unidirectional (following the wind direction), whereas impact ejecta often extends in multiple directions. Swiss cheese is a terrain type that consists of pits that are formed when the sun heats the ice making it sublimate (change solid to gas).

Other is a catch-all class that contains images that fit none of the defined classes of interest. This class makes up the majority of our data set.

Structural planes decrease the strength and stability of rock masses, severely affecting their mechanical properties and deformation and failure characteristics. Therefore, investigation and analysis of structural planes are crucial tasks in mining rock mechanics. The drilling camera obtains image information of deep structural planes of rock masses through high-definition camera methods, providing important data sources for the analysis of deep structural planes of rock masses. This paper addresses the problems of high workload, low efficiency, high subjectivity, and poor accuracy brought about by manual processing based on current borehole image analysis and conducts an intelligent segmentation study of borehole image structural planes based on the U2-Net network. By collecting data from 20 different borehole images in different lithological regions, a dataset consisting of 1,013 borehole images with structural plane type, lithology, and color was established. Data augmentation methods such as image flipping, color jittering, blurring, and mixup were applied to expand the dataset to 12,421 images, meeting the requirements for deep network training data. Based on the PyTorch deep learning framework, the initial U2-Net network weights were set, the learning rate was set to 0.001, the training batch was 4, and the Adam optimizer adaptively adjusted the learning rate during the training process. A dedicated network model for segmenting structural planes was obtained, and the model achieved a maximum F-measure value of 0.749 when the confidence threshold was set to 0.7, with an accuracy rate of up to 0.85 within the range of recall rate greater than 0.5. Overall, the model has high accuracy for segmenting structural planes and very low mean absolute error, indicating good segmentation accuracy and certain generalization of the network. The research method in this paper can serve as a reference for the study of intelligent identification of structural planes in borehole images.

The data originates from the Laboratory of Veterinary Medicine of the State University of Maranhão as part of a research project conducted at a Federal Institute of Maranhão. The histopathological image database was created to train CNN models for pet cancer classification. The image database was created from 82 original images that were enlarged using data augmentation such as vertical flip, horizontal flip, and random crop(px(0.50)), totaling 592 histopathological images divided into benign and malignant.

Training and test images of E. coli cells for artificial labeling of membranes in brightfield images using fnet or CARE, as well as trained models for prediction of super-resolution membranes.

Additional information can be found on this github wiki.

Example image shows an E. coli bright field image and PAINT membrane image predicted by the neural network (scale bar is 1 µm).

Training and testing dataset

Data type: Paired bright field and super-resolution images

Microscopy data type: Bright field and fluorescence microscopy (widefield and point accumulation for imaging in nanoscale topography (PAINT) images)

Microscope: Nikon Eclipse Ti-E equipped with an Apo TIRF 1.49NA 100x oil immersion objective

Cell type: E. coli K12 strain derivatives

File format: .tif (8-bit)

Image size: 512x512 px2 with different pixel sizes:

1x tube lens: 158 nm (raw) and 19.75 nm (8x upscaled for PAINT images)

1.5x tube lens: 106 nm (raw) (widefield fluorescence only)

fnet model (PAINT membrane images)

The fnet 2D model was generated using the ZeroCostDL4Mic platform (Chamier et al., 2021). It was trained for 200,000 steps on 33 paired images (image dimensions: (512 x 512 px²), patch size: (128 x 128 px²)) with a batch size of 4, a learning rate of 0.0004, 10% validation split and 4x data augmentation (flipping and rotation).

Model weights can be used with the ZeroCostDL4Mic fnet 2D notebook.

CARE model (PAINT membrane images):

The CARE 2D model was generated using the ZeroCostDL4Mic platform (Chamier et al., 2021). It was trained for 300 epochs (100 steps/epoch) on 33 paired images (image dimensions: 512 x 512 px², patch size: 256 x 256 px²) with a batch size of 4, a learning rate of 0.0004, 90/10% train/validation split and 4x data augmentation (flipping and rotation).

Model weights can be used with the ZeroCostDL4Mic CARE 2D notebook or the CSBDeep Fiji plugin.

Author(s): Christoph Spahn1,2, Mike Heilemann1,3

Contact email: christoph.spahn@mpi-marburg.mpg.de

Affiliation(s):

1) Institute of Physical and Theoretical Chemistry, Max-von-Laue Str. 7, Goethe-University Frankfurt, 60439 Frankfurt, Germany

2) ORCID: 0000-0001-9886-2263

3) ORCID: 0000-0002-9821-3578

The original PlantVillage dataset paper is An open access repository of images on plant health to enable the development of mobile disease diagnostics by Hughes & Salathe (2016). https://arxiv.org/abs/1511.08060

It was created with six different augmentation techniques for creating more diverse datasets with different background conditions. The augmentations used in this process were scaling, rotation, noise injection, gamma correction, image flipping, and PCA color augmentation.

PlantVillage Dataset Contributors - Arun Pandian J, Geetharamani Gopal

Description: The Anime Style Dataset is ideal for training AI models to perform style transformation between real human faces and anime-style illustrations. It contains two main folders:

👉 Download the dataset here Training Set: 820 pairs of real faces and their anime-style representations. Test Set: 93 similar pairs for validation. Applications: This dataset supports various AI research areas, including style transfer, GANs, and creative projects in animation and gaming industries.

Download Dataset Enhanced Features: Advanced Augmentation: To diversify the dataset, techniques such as flipping, rotation, or color changes can be applied. Metadata Enrichment: Including annotations for facial features, pose, and lighting conditions can help develop more nuanced models capable of capturing specific anime styles. Use Cases: Digital Content Creation: Automating the conversion of real-life photos into anime avatars for media or gaming. Artistic Tools: Enabling new tools for anime artists by simplifying style transition between real faces and animated ones. AI-driven Filters and Augmented Reality: Expanding its use in creative social media filters and personalized avatars for various platforms. This dataset is sourced from

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