- A Conditional Generative Network for synthesizing Fluroscein Angio images from Fundus Photography.
- Automated denoising of calcium signals using mathematical noise modeling.
- Incorporating Supervised-Unsupervised Joint Attention Network for improving robustness for recogniziing diabetic Retinopathy.
- A model for grading different Retinal Diseases using Deep Learning from Optical Coherence Tomography Images.
- A novel deep residual network for state-of-the-art traffic sign recogntition for German and Belgian Traffic Sign data-set.
- A Fully Convolutional Neural Network with multi-scale skip connection and dilated convolution for Semantic Segmentation.