26 datasets found

The AVA v2.2 dataset for spatiotemporal action localization contains the bounding box annotations and the corresponding action labels on keyframes.

The Diving48 dataset is a fine-grained video dataset of competitive diving. It has ∼18k trimmed video clips of 48 unambiguous dive sequences standardized by the professional....

The Something-Something dataset consists of 174 fine-grained action categories that depict humans performing everyday actions with common objects. Recognizing actions in the...

The Something-Something-V2 dataset is a large-scale video action recognition dataset.

The Something-Something v2 (SSv2) dataset is a large collection of video clips of humans performing actions with everyday objects.

This work explores the capability of a simple baseline called ViFi-CLIP (Video Fine-tuned CLIP) for adapting image-based CLIP to video domain.

The Mini-Kinetics dataset is a mini version of the Kinetics-400 dataset, containing 240k training samples and 20k validation samples in 400 human action classes.

The HMDB51 dataset is a video dataset for human action recognition. It contains 6,767 videos annotated with 51 categories of human actions.

The Kinetics-700 dataset is a large-scale video dataset for human action recognition. It contains 555,774 videos annotated with 700 categories of human actions.

Video classification is a fundamental problem in many video-based tasks. Applications such as autonomous driving technology, controlling drones and robots are driving the demand...

The dataset used in the paper for self-supervised learning of video representations.

The EGTEA Gaze+ dataset offers approximately 10,000 samples of 106 non-scripted daily activities that occur in a kitchen.

The dataset used in the paper is HMDB51 and UCF101.

The dataset used in the paper is Kinetics-400 and Something-Something-V2.

The Kinetics dataset is a large-scale action recognition dataset. It contains videos of various actions performed by humans, with annotations of the actions performed.

UCF-101 dataset is a large-scale action recognition dataset, containing 13,320 videos categorized into 101 human action categories.

The Kinetics-400 and Kinetics-600 datasets are video understanding datasets used for learning rich and multi-scale spatiotemporal semantics from high-dimensional videos.

The Kinetics-600 dataset consists of 392k training videos and 30k validation videos in 600 human action categories.

Motion has shown to be useful for video understanding, where motion is typically represented by optical flow. However, computing flow from video frames is very time-consuming....

Motion has shown to be useful for video understanding, where motion is typically represented by optical flow. However, computing flow from video frames is very time-consuming....