Aryan Karn

I am a 3rd year undergraduate student at Motilal Nehru National Institite of Technology Allahabad, India where I study Electronics and Communicition Engineering. I am specializing in the field of Artificial Intelligence.

I have been extremely fortunate to be able to voluntare for RED CROSS NEPAL and also intern for Agriculture Development Bank, Nepal. Currently I am doing a research intern at Institute of Engineering, IOE, Pulchowk Engineering Campus on model-based reinforcement learning inspired by cognitive theories.

I am interested in the symbiotic relationship between AI and cognitive science i.e. AI to better understand the brain and cognitive theories to help build better AI models.

Email  /  CV  /  GitHub  /  LinkedIn  /  Instagram

• Sep 15, 2020: I have started preparing for the GRE
• Sep 5, 2020: I have been working on Word2Brain2Image research project.
• Aug 6, 2020: I will be attending the MIT Brain, Minds, and Machines Summer Course starting from August 10th!
• Jun 18, 2020: I was the lead speaker at AI Nepal.org held virtually on TF.Keras and Image Classification.
• May 5, 2020: I have been selected to work on covid19.mohp.gov.np for the Morang District.

In this project, the main motivation was to infer about a person’s look from the way they speak. We design and train a deep neural network to perform this task using thousands of natural YouTube videos of people speaking.Creating a model for reconstructing a person's face from his/her voice sample.

Parameter sharing is the major reason of success of building large models for deep neural networks. The paper proposed by Shuangfei Zhai, Yu Cheng, Weining Lu and Zhongfei Zhang (NIPS 2016) introduces the idea of Doubly Convolutional Neural Networks, which significantly improves the performance of CNN with the same number of parameters. We have implemented DCNN as a part of Statistical Methods in AI Project.

• Organized and managed food pantry operations resulting in a 10% decrease in spending.
• Worked in the Remote District of Nepal to and taught childrens studying in primary school.
• Helped in spreading the awearness of using Digital Payment in Rural District of Nepal.

Revamped the existing Information Retrieval system to focus more on distributional semantics. Also worked on the ADBL Bank online Banking App for the Document scaning computer vision process.

Working on human-level reinforcement learning inspired by intuitive cognitive theories to mimic human behavior on complex sequential tasks.

NumPy is a library for the Python programming language, adding support for large, multi-dimensional arrays and matrices, along with a large collection of high-level mathematical functions to operate on these arrays.

Image captioning with vanilla RNNs and LSTMs on the COCO dataset. Network visualizations and style transfer on SqueezeNet, which is a deep CNN pretrained to perform image classification on ImageNet.

In this project, I am generating human faces which probably does not exist in real life. I will be using the generative adversarial network(GAN) for the task. I am using CelebA dataset for training the network. This dataset contains 2,00,000 images of well-known people.

I've implemented the Vehicle Detection using Mask R-CNN and Computer Vision based Lane Detectionusing on Keras and TensorFlow. The model detects vehicles in the image frame using segmentation masks with the pretrained weights trained on COCO dataset; the lane detection is done using sobel filter.

An open source implementation of Machine Translation english to french using Seq2Seq Attention model in PyTorch.

An open source implementation of ChronoNet.

Research Project in unofficial collaboration with TCS Research. Applying State of the art models for pneumonia detection on RSNA pneumonia detection dataset. Tested InceptionNet-v3, DenseNet121 and explored Mask RCNN applicability for the dataset. Got 83.8% and 77.9% classification accuracy respectively.

Template