Teaching the Donkey car to drive a track in the simulator using State Representation Learning and different Reinforcement Learning Algorithms including Deep Q-Network, Soft Actor-Critic and Proximal Policy Optimization Algorithms.

Source code and data for paper ``Unsupervised Dual Paraphrasing for Two-stage Semantic Parsing" in ACL 2020.

Collection of autoencoder models in Tensorflow

Autoencoders test from Coursera's Advanced Machine Learning - Intro to Deep Learning course.

Autoencoders with TensorFlow 2

Implementation and evaluation of classical and deep learning-based image denoising methods. The project compares Non-Local Means (NLM), BM3D, U-Net, and Denoising Autoencoders under different noise models (Gaussian, Salt & Pepper), using PSNR as the primary performance metric.

PyTorch DAE that denoises CIFAR-10 images — 24.62 dB PSNR, 0.8225 SSIM, 182K params. Features 45+ publication-quality figures, noise-level & bottleneck ablation studies, and a full 10-page LNCS report.

A research-grade PyTorch framework for robust object recognition under extreme environmental noise. Implements self-supervised Denoising Autoencoders (DAE) with ResNet/ViT architectures on the official CIFAR-10-C benchmark. Includes Grad-CAM interpretability and automated robustness benchmarking.