Abstract
This work studies large scale visual recognition, specifically dense image captioning, using deep learning on GPUs and Apache Spark. It covers an efficient implementation of convolutional and recurrent neural networks in Apache Spark, and an implementation of neural style transfer that uses a convolutional neural network to combine the content and style of different images.
B.Tech thesis, Department of Computer Science and Engineering. Supervised by Dr. Sourangshu Bhattacharya.
