Pikabot-RL
Using RL for training bots to play the online battle game Pokémon Showdown
Diffusion Everything
Demos involving diffusion models : 1. Training and inference of diffusion on a custom 2D dataset. 2. Training and inference of VAE and latent diffusion models 3. Inference of large diffusion models with different reverse samplers
Segment-then-Edit
Segment parts of image and then perform inpainting or editing on specific object.
Image-Alchemy
A two-stage personalization pipeline for personalized image generation using LoRA-based attention fine-tuning and segmentation-guided Img2Img synthesis.
Knowledge Graph Embeddings for VLMs
Analysing impact of vector and graph embeddings on retrieval generation of VLMs on downstream tasks.
FeedCode
FeedCode is an intelligent feedback tool that provides personalized code reviews based on your coding style and activity
Visual ML
Interactive Visual Machine Learning Demos. Machine Learning in the browser powered by TF JS.
d2l-pytorch
This project reproduces the book Dive Into Deep Learning (www.d2l.ai), adapting the code from MXNet into PyTorch.
Graph Representation Learning
Implementation and Explanation Graph Representation Learning papers involving DeepWalk, GCN, GraphSAGE, ChebNet and GAT.
Reinforcement Learning 2048
This project builds on a demo for playing the game 2048 using Reinforcement Learning.
Sarcasm Detection using BERT
Implementation of BERT based Sarcasm Detection Classification model using Tensorflow.
Adversarial Lab
This project builds on a demo for several Adversarial Attacks on ImageNet Classifier Models.
Number Plate Detection
This is a step towards automating the vehicles entering IIT Roorkee campus using state of the art Deep learning & Computer Vision.
Eye In The Sky: Image Segmentation Challenge Inter IIT 2018
Winning Solution in the Inter IIT Tech Meet 2018 Machine Learning Challenge, "Eye In The Sky".
Visualizing Loss Functions
Visualising different loss and optimisation functions using Autoencoders.
Neural Style Transfer
Implementation of Neural Style Transfer described in the paper by Gatys et.al.
Beginner Friendly Project: Traffic Sign Classification
Using GTSRB dataset to build a CNN based Traffic Sign Classifier.