Mehmet OKUYAR's Projects
Fast image augmentation library and easy to use wrapper around other libraries. Documentation: https://albumentations.ai/docs/ Paper about library: https://www.mdpi.com/2078-2489/11/2/125
This is Andrew NG Coursera Handwritten Notes.
Augmentation yolo format txt and images
This function I wrote performs the image and txt saving process. You can also use this function on a model you have trained before. If your model works well enough, it will easily detect the images and record the labels. You can edit the recorded data and train with your new data quickly.
Data is a huge factor in deep learning algorithms. The larger our data size, the better our model can generalize and learn. However, data preparation is a very laborious and time-consuming process. That's why I wanted to develop an application that I thought would make this stage easier. By using image processing techniques, it can track an object of your choice to a certain extent and saves the image and .txt file to the folder during tracking. Currently, it only works for one class and you can only label one object.
If you are developing a segmentation model, you may need to convert your tags from json format to mask format. Thanks to the interface we have developed, you will be able to easily turn your labels into masks. Also, if you have masks and want to convert them to json format, you can easily do this with our application. This conversion comes to you in Mask R-CNN format, so if you have mask photos, you can easily convert them to .json files suitable for Mask R-CNN format. This convenience will allow you to convert the labels you have as you wish, even if they do not fit your format.
You can use dcim images to convert them to png format
YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )
YOLO ROS: Real-Time Object Detection for ROS
Determination of drivable area in highway images with semantic segmentation.
Code and data for our paper "High-Fidelity 3D Digital Human Creation from RGB-D Selfies".
This repository contains python notebook for generating new set of images from existing images using Generative Adversarial Networks. The code has been tested on MNIST Dataset and can be extended to any other dataset
Redesigned Instagram application that works in real time using the Instagram API
A model on lane detection has been developed with the data set we have created.
Human Activity Recognition example using TensorFlow on smartphone sensors dataset and an LSTM RNN. Classifying the type of movement amongst six activity categories - Guillaume Chevalier
My personal repository
Official implementation of OOTDiffusion: Outfitting Fusion based Latent Diffusion for Controllable Virtual Try-on
Easy to generate user data
RobustSAM: Segment Anything Robustly on Degraded Images (CVPR 2024)
A lightweight vision library for performing large scale object detection/ instance segmentation.
We developed a python UI based on labelme and segment-anything for pixel-level annotation. It support multiple masks generation by SAM(box/point prompt), efficient polygon modification and category record. We will add more features (such as incorporating CLIP-based methods for category proposal and VOS methods for video datasets
A repository contains the code for various semantic segmentation in TensorFlow and PyTorch framework.
[CVPR2024] Official implementation of SplattingAvatar.
Spot SDK repo