20 datasets found

An audio-visual dataset of five crowded scenes: 'Riot', 'Noise-Street', 'Firework-Event', 'Music-Event', and 'Sport-Atmosphere'.

The CORSMAL Containers Manipulation dataset is a dataset of audio-visual recordings of people interacting with containers.

The dataset is used for learning general and effective models for both audio and video analysis from self-supervised temporal synchronization.

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

The LRS2 dataset consists of 48,164 video clips from outdoor shows on BBC television. Each video is accompanied by an audio corresponding to a sentence with up to 100 characters.

The LRS3 dataset is a large-scale dataset for visual speech recognition. It consists of thousands of spoken sentences from TED videos.

The LRS2 dataset consists of news recordings from the BBC, with different lighting, backgrounds, face poses, and people with different origins.

The GRID dataset was introduced by [5] as a corpus for tasks such as speech perception and speech recognition. GRID contains 33 unique speakers, articulating 1000 word sequences...

Estimating visual information from audio through manifold learning.

Visual to sound: Generating natural sound for videos in the wild.

Deep cross-modal audio-visual generation.

Sound2Scene is a sound-to-image generative model and training procedure that addresses the challenges of dealing with the large gaps that often exist between sight and sound.

The AVQA dataset contains 57,015 videos and 57,335 question-and-answer pairs.

The Music-AVQA dataset contains multiple question-and-answer pairs, with 9,288 videos and 45,867 question-and-answer pairs.

Audio-visual question answering (AVQA) requires reference to video content and auditory information, followed by correlating the question to predict the most precise answer.

A dataset of audio-visual pairs where the audio is visually indicated.

A large-scale audio-visual dataset containing audio-visual pairs.

The dataset used in the paper for 3D head avatar reconstruction from monocular RGB videos.

The CREMA-D dataset is an audio-visual dataset for emotion recognition task, each video in which consists of both facial and acoustic emotional expressions.

The MEAD dataset is a large-scale, high-quality emotional audio-visual dataset, which consists of 60 actors, including 8 basic emotions and 3 different emotional-intensity...