Published May 26, 2017

Unsupervised Depth Perception

Tinghui Zhou unveils his pioneering research on unsupervised depth perception and its transformative implications for autonomous driving and digital media, revealing how unlabeled video data can teach computers geometry akin to human visual learning.

Episode Highlights

Topics covered

Episode Highlights

Related Episodes

Visualization and Interpretability
Answers 383 questions
Neuroscience from a Data Scientist's Perspective
Answers 383 questions
Fooling Computer Vision
Answers 383 questions
Understanding Neural Networks
Answers 383 questions
Video Anomaly Detection
Answers 383 questions
Visual Illusions Deceiving Neural Networks
Answers 383 questions
[MINI] The Perceptron
Answers 383 questions
Uncertainty Representations
Answers 383 questions
Easily Fooling Deep Neural Networks
Answers 383 questions
k-means Image Segmentation
Answers 383 questions
Machine Learning Done Wrong
Answers 383 questions
The Experimental Design of Paranormal Claims
Answers 383 questions
Blind Spots in Reinforcement Learning
Answers 383 questions
Detecting Terrorists with Facial Recognition?
Answers 383 questions
Introduction
Answers 383 questions

Dexa/Kyle Polich

Unsupervised Depth Perception

Topics covered

Popular Clips

Exploring Data Research

Advancements in Depth Estimation

Tackling Occlusion

Visual Perception Illusions

Explaining Explainability Mask

Predicting Frames, Moving Objects

Deep Learning Architecture

Advancing Self-Driving Cars

Visualizing Data

Unsupervised Learning

Episode Highlights

Depth Perception

Applications in Self-Driving

Unsupervised Learning

Related Episodes

Visualization and Interpretability

Neuroscience from a Data Scientist's Perspective

Fooling Computer Vision

Understanding Neural Networks

Video Anomaly Detection

Visual Illusions Deceiving Neural Networks

[MINI] The Perceptron

Uncertainty Representations

Easily Fooling Deep Neural Networks

k-means Image Segmentation

Machine Learning Done Wrong

The Experimental Design of Paranormal Claims

Blind Spots in Reinforcement Learning

Detecting Terrorists with Facial Recognition?

Introduction

Unsupervised Depth Perception

Topics covered

Popular Clips

Episode Highlights

Depth PerceptionTinghui Zhou discusses his groundbreaking work on unsupervised learning for depth and ego-motion from video. His research focuses on transforming 2D data into 3D visualizations, with applications in fields like autonomous driving.

Depth Perception

Applications in Self-DrivingTinghui Zhou discusses the potential of his research to transform self-driving technology and digital media. His work on unsupervised depth perception could enhance geometry and motion estimation, offering new possibilities for autonomous vehicles and video applications.

Applications in Self-Driving

Unsupervised LearningThe episode explores the innovative use of unsupervised methods for depth perception, as explained by Tinghui Zhou. He discusses how their approach leverages unlabeled video data to teach computers geometry, mimicking human visual learning.

Unsupervised Learning

Related Episodes