EmotionAI: Facial Key Point Detection
A deep learning model based Convolutional Neural Network with residual blocks to predict facial key points; a fundamental step in emotion detection from facial images.
Projects
Data Mining Projects
I tried to implemented two Data Mining theories in Object Detection Problem. One of them is Object Detection with descriptive statistical feature, and the other one is with Decision Tree.
Brain Tumour auto-segmentation from 3D MRI images using 3D U-Net
Building a neural auto-segmentation model for MRI images with UNets with Dice Loss function. In this project, I build a multi-class segmentation model. We’ll identify 3 different abnormalities in each image: edemas, non-enhancing tumors, and enhancing tumors. This project is to be continued.
Bachelor Thesis: Automated method to segment retinal blood vessels from color fundus photographs.
In this work, we propose an innovative approach to segment the retinal blood vessel in the presence of pathology. The method combines both supervised and unsupervised approaches in the retinal imaging context. Experiments are conducted on three publicly available datasets named: DRIVE, STARE, and CHASE DB1, and the proposed method has been compared against the state-of-the-art methods. The proposed method is found about 1% more accurate than the best performing supervised method and 2% more accurate than the state-of-the-art Nguyen et al.’s method.
Local Haar Pattern: A Feature Descriptor for Biomedical Images
In order to compute the LHP descriptor a patch p of size 32×32 is consider around the pixel of interest, and vector of size 128 bytes is calculated that represents the patch. Each byte of the vector is computed based on the intensity comparisons of two-pixel groups.