Smart Music Solution, based on Tencent Multimedia Lab's proprietary music understanding technology, helps users deeply understand, analyze, and create music, explore more music scenarios, and discover music value.
Capability Overview
Music Understanding and Processing
Understand, analyze, and process music content and structure, including the following capabilities:
- Music Beat Recognition: Input original music, automatically locate drum beat positions/time points
- Music Highlight Recognition: Input original music, detect highlight events, automatically locate all chorus sections
- Skip Music Intro: Input original music, automatically identify the time point when the intro ends (singing starts)
- Source Separation: Input original music, separate each track, output vocals, accompaniment, drums, bass, and other 4 individual tracks
- Vocal Score: Score the music to be evaluated based on reference music across various dimensions such as pitch and rhythm
- Music Tag: Input original music, understand music's intensity, emotion, style, sentiment, and other dimensions, automatically generate and output music tag results for each system
- Vocal to MIDI: Input original music, separate the vocal part, calculate the pitch and range of each note, output standard MIDI file
Intelligent Music Generation
Intelligently synthesize, create, and adapt singing or accompaniment, including the following capabilities:
- Intelligent Composition: Based on a few notes or conditional prompts, automatically create single-track melodies or multi-track accompaniment (Coming Soon)
- Singing Synthesis: Automatically synthesize high-fidelity singing dry vocals from sheet music (Coming Soon)
Music Cross-Modal Fusion
Deeply integrate music with cross-modal information such as video, text, and images to complete intelligent content production, including the following capabilities:
- Short MV: Input original MV, automatically edit highlight MV clips from spatial and temporal dimensions, and add the highlight part of the original music
- Video BGM: Analyze video scenes, objects, events, and other information, recommend music with matching style, emotion, and rhythm from the music library to synthesize background music (Coming Soon)
- Music to Dance: Generate matching dance movements based on music melody, rhythm, and emotion, supporting coordinate information and animation video output (Coming Soon)
