1. Emotion Recognition algorithm development

1. Developed Traffic Sign Recogntion algorithm for German signs.

2. Utilized computer vision based algorithms, Deep learning models(SSD, YOLO) and Machine learning algorithms(SVM, Random Forest, Logistic Regression)

Acceleration of Image processing algorithms on GPU

Optimization of Lucas Kanade algorithm

1 Develop traffic sign detection algorithm for autonomous cars using Machine learning and deep learning. The development includes the detection of traffic sign and gets it trained through classifiers and deployed it on vehicle. 2 Developed algorithm for emergency vehicle light detection based on computer vision. 3 Validating the influence of lossy compression in object detection for autonomous driving application using deep learning.

Accelerate image processing algorithms for medical applications and evaluate its performance on GPU.

To evaluate the performance of Artificial Neural network algorithms, on GPU to achieve the maximum performance.

The project is about optimization of the given image processing algorithms using DMIP library for the monochrome and RGB printer pipeline. The given image processing algorithms are ported and optimized using DMIP libraries in a specified platform(Sandy Bridge) and the performance of the optimized code is measured.

Optimization of Lucas Kanade feature tracking algorithm using SSE, AVX intrinsic on Intel platform.

Development of image processing algorithm for video surveillance camera application.

Optimization of H.264 Video Encoder.