79 datasets found

The proposed network is implemented in PyTorch using a single Tesla V100-DGXS-32GB GPU of an NVIDIA DGX station. For the ResNet-50 part, PyTorch default ImageNet pretrained...

Face recognition and person re-identification using paired image-attribute data, where the attributes (i.e., soft biometrics) are only available during the training phase.

THUMOS Challenge: Action Recognition with a Large Number of Classes.

Generating human action proposals in untrimmed videos is an important yet challenging task with wide applications. Current methods often suffer from the noisy boundary locations...

The VGGSound dataset is a large-scale audio-visual dataset containing 10,000 10-second video clips with corresponding audio files.

The dataset contains videos of people performing normal activities and falls.

The dataset contains videos of people performing normal activities and falls.

The proposed spatio-temporal adversarial learning framework for detecting unseen falls from videos.

The SEWA dataset contains video clips annotated with facial landmarks, valence, and arousal.

The AFEW-VA dataset contains video clips annotated with valence and arousal.

Temporal action proposal generation is an important task, akin to object proposals, temporal action proposals are intended to capture “clips” or temporal intervals in videos...

The MuscleMap dataset is a large-scale video-based dataset for Activated Muscle Group Estimation (AMGE) in the wild task. It contains 15,004 video clips with 135 different...

MoVi is a large multi-purpose human motion and video dataset.

BMLrub is a dataset of human motion and video.

The dataset used in the paper is not explicitly described, but it is mentioned that the authors tested their model on various signal reconstruction tasks: 1D sinusoidal...

The CelebV-HQ dataset contains 35,666 face videos from over 15,000 identities.

The UCF-Crime and XD-Violence datasets are used for weakly supervised video anomaly detection.

Weakly-Supervised Temporal Action Localization (WSTAL) aims to localize actions in untrimmed videos with only video-level labels.

CamVid dataset is a benchmark dataset for semantic segmentation. It consists of 700 images with 11 object classes.

BDD100K Dataset is a large-scale dataset for autonomous driving, containing 100,000 images, with 20,000 images for training and 80,000 images for testing.