A Sparsely Labeled Dataset for Action Classification and Localization - SLAC

Explore a comprehensive overview of the Sparsely Labeled Dataset for Action Classification and Localization (SLAC) in this 36-minute lecture from the University of Central Florida. Delve into the motivation behind the dataset's creation, understand the data collection pipeline, and learn about the various classifiers used. Examine the sampling distribution, quality control measures, and validation processes employed. Compare SLAC with other datasets and gain insights into its evaluation methods for human action detection and classification in videos.

Introduction
Motivation
Datasets
Datasets for Video Class
Data Collection Pipeline
Training and Validation
Negative Samples
Internal Deduplication
Human Action Detection
Flowchart
Classifiers
YouTube Frame Based
Google Image Based
Sampling Distribution
Quality Control
Validation
Comparison
Evaluation