Tuhin Mukherjee's repositories
elysium
Z- algorithm for pattern matching, Trie-Aho-Corasick Automaton(FGREP), Hungarian Algorithm for the Munkres Assignment Problem, Binary Lifting, Eulerian Tour for Least Common Ancestor(LCA) using Sparse Table coupled with Farach-Colton and Bender optimization, Wellsh Powell Algorithm for Graph Coloring, Kahn's Agorithm for TopSort and Cycle detection in Graphs. Sparse Table for range queries. Johnson's Algorithm. AHU Encoding. Kosaraju's Algorithm for strongly connected components. Tarjan's Algorithm. Union Find. Fenwick Tree. Segment Tree. Ford Fulkerson. Edmonds-Karp. Dinic's Algorithm. Capacity Scalling. Push Relabel. Dijkstra's Algorithm. Bellman Ford's Algorithm. Floyd Warshall's Algorithm and much more..
Convolutional-AutoEncoder-for-Image-Compression-and-Denoising-PCA
A convolutional autoencoder to solve the problem of image compression and image denoising at a much lower bottlneck value in terms of storage. A little finetuning would enable the model to solve the problem of Image inpainting as well. Added demonstration of Principal Component Analysis as welll.
Java-Swing-Covid-19-Simulator-Live-Tracing
A Covid-19 simulator built using the Java Swing module with added features like live tracing. A probabilistic virtual simulation for real-life data estimation.
Core-Python-Character-level-RNN-for-Text-generation
A character level Recurrent Neural Network for text generation coded using raw Python without using any deep learning framework or dependency except Numpy.
Coursera-MultiLabel-Classification-using-Binary-CrossEntropy-on-the-IMDB-dataset-OVRC
A Coursera assignment for MultiLabel Classification using a One vs Rest (OVRC) approach.
Java-Google-Foobar-Running-With-Bunnies
An application of Floyd-Warshall's Algorithm, Johnson's Algorithm for weight readjustment, Bellman-Ford's Algorithm and DFS coupled with memoization.
Python-Chatroom-EventLoop
This is a demonstration of Asynchronous Non-Blocking threads and callbacks through the simulation of a Reactor EventLoop without using asyncio. This is demonstrated through the working of a chatroom that is monitored by a server and accessed by clients. The reactor eventloop could very well be replaced by a Proactor eventloop as well.
Pytorch-Fashion-MNIST
Fashion-MNIST classification using PyTorch.
Pytorch-Image-Colourisation-using-RGB-predictor
A finetuned U-Net model trained on the Modified Celeb-A dataset(containing face masks) for solving the problem of image colorisation. The model outputs normalized RGB values which is in contrast to the benchmark model which outputs LAB values.
PyTorch-U-Net-for-Semantic-Segmentation
An implementation of the U-Net model based on the paper by Olaf Ronneberger trained on the Pascal-VOC dataset
Tensorflow-Keras-CycleGAN
A CycleGAN implementation using Tensorflow :: Keras
Tensorflow-Keras-Semantic-Segmentation
A self-designed model to solve the computer vision problem of Semantic Segmentation trained on the Pascal-VOC datastet using the Keras API of the Tensorflow framework in Python.
C-Leetcode-1263-Hard-Minimum-Ways-to-push-a-box-to-a-target
Using a structured BFS tree implementation in C
Game-Demo-UI
This is a very basic game UI built using Java Media Framework(JMF) and ASCII-ART with added python utility scripts for screen recording and video trimming.
Proto
A core JAVA implementation of the Sliding Window Protocols, Go-Back-N Automatic Repeat Request(ARQ) and Selective Repeat Automatic Repeat Request(ARQ), along with the Stop-And-Wait ARQ included.