Self-supervised Learning for Autonomous Vehicles

This lecture explores the concept of self-supervised learning for autonomous vehicles, focusing on deriving labels from data itself rather than human annotation. It discusses the limitations of deep supervised learning, the success of self-supervised learning, and the application of techniques like Word2Vec. The lecture delves into various self-supervised tasks such as denoising autoencoders, in-painting, and colorization, highlighting their importance in learning powerful representations from unlabeled data. It also covers topics like social contrastive learning and instance discrimination, showcasing their role in vision tasks. The lecture concludes by discussing challenges, test-time adaptation, and the potential of self-supervised learning to address distributional shift in autonomous vehicle applications.

About this result

This page is automatically generated and may contain information that is not correct, complete, up-to-date, or relevant to your search query. The same applies to every other page on this website. Please make sure to verify the information with EPFL's official sources.

Related lectures (156)

Related concepts (46)

CIVIL-459: Deep learning for autonomous vehicles

Deep Learning (DL) is the subset of Machine learning reshaping the future of transportation and mobility. In this class, we will show how DL can be used to teach autonomous vehicles to detect objects,

Lethal autonomous weapon

Lethal autonomous weapons (LAWs) are a type of autonomous military system that can independently search for and engage targets based on programmed constraints and descriptions. LAWs are also known as lethal autonomous weapon systems (LAWS), autonomous weapon systems (AWS), robotic weapons or killer robots. LAWs may operate in the air, on land, on water, underwater, or in space. The autonomy of current systems was restricted in the sense that a human gives the final command to attack—though there are exceptions with certain "defensive" systems.

Autonomous robot

An autonomous robot is a robot that acts without recourse to human control. The first autonomous robots environment were known as Elmer and Elsie, which were constructed in the late 1940s by W. Grey Walter. They were the first robots in history that were programmed to "think" the way biological brains do and meant to have free will. Elmer and Elsie were often labeled as tortoises because of how they were shaped and the manner in which they moved. They were capable of phototaxis which is the movement that occurs in response to light stimulus.

Vehicular automation

Vehicular automation involves the use of mechatronics, artificial intelligence, and multi-agent systems to assist the operator of a vehicle (car, aircraft, watercraft, or otherwise). These features and the vehicles employing them may be labeled as intelligent or smart. A vehicle using automation for difficult tasks, especially navigation, to ease but not entirely replace human input, may be referred to as semi-autonomous, whereas a vehicle relying solely on automation is called robotic or autonomous.

Uncrewed vehicle

An uncrewed vehicle or unmanned vehicle is a vehicle without a person on board. Uncrewed vehicles can either be under telerobotic control—remote controlled or remote guided vehicles—or they can be autonomously controlled—autonomous vehicles—which are capable of sensing their environment and navigating on their own. There are different types of uncrewed vehicles: Remote control vehicle (RC), such as radio-controlled cars or radio-controlled aircraft Unmanned ground vehicle (UGV), such as the autonomous cars

Artificial neural network

Artificial neural networks (ANNs, also shortened to neural networks (NNs) or neural nets) are a branch of machine learning models that are built using principles of neuronal organization discovered by connectionism in the biological neural networks constituting animal brains. An ANN is based on a collection of connected units or nodes called artificial neurons, which loosely model the neurons in a biological brain. Each connection, like the synapses in a biological brain, can transmit a signal to other neurons.

Trajectory Forecasting in Autonomous Vehicles CIVIL-459: Deep learning for autonomous vehicles

Explores trajectory forecasting in autonomous vehicles, focusing on deep learning models for predicting human trajectories in socially-aware transportation scenarios.

Neural Networks Optimization CIVIL-459: Deep learning for autonomous vehicles

Explores neural networks optimization, including backpropagation, batch normalization, weight initialization, and hyperparameter search strategies.

Generative Models: Trajectory Forecasting CIVIL-459: Deep learning for autonomous vehicles

Explores generative models for trajectory forecasting in autonomous vehicles, including discriminative vs generative models, VAES, GANS, and case studies.

Autonomous Vehicles: Trajectory Prediction and Social Behavior CIVIL-459: Deep learning for autonomous vehicles

Explores challenges in deep learning for autonomous vehicles, emphasizing social behavior modeling and feasible trajectory prediction.

Deep Learning for Autonomous Vehicles: Predictive Models CIVIL-459: Deep learning for autonomous vehicles

Explores predictive models and trackers for autonomous vehicles, covering object detection, tracking challenges, neural network-based tracking, and 3D pedestrian localization.