Ádám Divák's repositories
vgg_deconv_vis
Visualising what a convolutional neural network 'sees' using the Deconvnet technique, which identifies parts of an image that a given neuron/layer is sensitive to. Analysis performed using VGG.
udacity_sd_lidar_fusion
Lidar 3D object detection using NNs. Lidar-camera fusion using an EKF model, data association and track management
time_interpret
Code for On the reproducibility of: "Learning Perturbation to Explain Time Series Predictions", based on Joseph Enguehard's time series interpretability library
udacity_sd_control
Control and trajectory tracking for AVs tested in Carla. Implementing separate PID controllers for throttle and steering, controller parameter tuning using the twiddle algorithm, and some fixes to the original planner and simulator client
diffusion_augmented_pixelsplat
Improving novel view synthesis of 3D Gaussian splats using 2D image enhancement methods
equivariant_transfer_learning
Reproduction and extension of "Efficient Equivariant Transfer Learning from Pretrained Models" by Basu et al. (2023)
udacity_sd_planning
Basic behavior and motion planning tested in Carla. State machine-based behavior planning, motion planning using cubic spirals, velocity profile generation and cost-function based trajectory selection with static collision checking
udacity_sd_advanced_lanes_cv
Lane detection using traditional CV algorithms - calibration, birds-eye view perspective transformation, sliding windows based lane finding and temporal tracking
udacity_sd_basic_lanes_cv
Basic lane detection using traditional CV algorithms. Thresholding, Canny edge detection, line clustering using DBSCAN, ego lane separator selection and visualization
udacity_sd_traffic_sign_classification
Traffic sign classification based on the LeNet architecture with dropout, regularization and hyperparameter optimization.
udacity_sd_vision_nn
Starter Code for the Course 1 project of the Udacity Self-Driving Car Engineer Nanodegree Program
udacity_sd_vision_nn_aws
Object Detection in an Urban Environment using TF2 Object Detection API and AWS Sagemaker.