Foley Music: Learning to Generate Music from Videos
This addresses the challenge of creating synchronized and editable music for video content, though it is incremental as it builds on motion-to-audio translation with a focus on interpretability.
The paper tackles the problem of generating music from silent videos of people playing instruments by translating body movements into MIDI events, resulting in a system that outperforms existing methods in producing pleasant music.
In this paper, we introduce Foley Music, a system that can synthesize plausible music for a silent video clip about people playing musical instruments. We first identify two key intermediate representations for a successful video to music generator: body keypoints from videos and MIDI events from audio recordings. We then formulate music generation from videos as a motion-to-MIDI translation problem. We present a Graph$-$Transformer framework that can accurately predict MIDI event sequences in accordance with the body movements. The MIDI event can then be converted to realistic music using an off-the-shelf music synthesizer tool. We demonstrate the effectiveness of our models on videos containing a variety of music performances. Experimental results show that our model outperforms several existing systems in generating music that is pleasant to listen to. More importantly, the MIDI representations are fully interpretable and transparent, thus enabling us to perform music editing flexibly. We encourage the readers to watch the demo video with audio turned on to experience the results.