LoVA

Demos

We select both 10-second samples from VGGSound and long-form samples form UnAV100. Since the video is relatively long, we appreciate your patience while the server loads the content. Thank you!
It is recommended to use earphones to hear the demos videos, raise the volume and zoom in the videos.

Section 1: Long-Form UnAV100 results.

Sample 1: Playing the flute.

Ground Truth	SpecVQGAN	IM2WAV	Diff-Foley

	foleycrafter	TiVA	LoVA (Ours)

Sample 2: Wind, wave, boat.

Ground Truth	SpecVQGAN	IM2WAV	Diff-Foley

	foleycrafter	TiVA	LoVA (Ours)

Sample 3: Cattle mooing.

Ground Truth	SpecVQGAN	IM2WAV	Diff-Foley

	foleycrafter	TiVA	LoVA (Ours)

Sample 4: Raining, lighting and thunder.

Ground Truth	SpecVQGAN	IM2WAV	Diff-Foley

	foleycrafter	TiVA	LoVA (Ours)

Sample 5: Playing ping pong.

Ground Truth	SpecVQGAN	IM2WAV	Diff-Foley

	foleycrafter	TiVA	LoVA (Ours)

Section 2: 10-second VGGSound results.

Sample 1: Footstep, ice breaking.

Ground Truth	SpecVQGAN	IM2WAV	Diff-Foley

	foleycrafter	TiVA	LoVA (Ours)

Sample 2: Engine starting.

Ground Truth	SpecVQGAN	IM2WAV	Diff-Foley

	foleycrafter	TiVA	LoVA (Ours)

Sample 3: Child groaning.

Ground Truth	SpecVQGAN	IM2WAV	Diff-Foley

	foleycrafter	TiVA	LoVA (Ours)

Sample 4: Owl hooting.

Ground Truth	SpecVQGAN	IM2WAV	Diff-Foley

	foleycrafter	TiVA	LoVA (Ours)

Sample 5: Playing the violin.

Ground Truth	SpecVQGAN	IM2WAV	Diff-Foley

	foleycrafter	TiVA	LoVA (Ours)