SOTAVerified

Audio Generation

Audio generation (synthesis) is the task of generating raw audio such as speech.

( Image credit: MelNet )

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 76–100 of 270 papers

TitleStatusHype
MusicLDM: Enhancing Novelty in Text-to-Music Generation Using Beat-Synchronous Mixup StrategiesCode1
Make-An-Audio 2: Temporal-Enhanced Text-to-Audio GenerationCode1
An Efficient Membership Inference Attack for the Diffusion Model by Proximal InitializationCode1
Any-to-Any Generation via Composable DiffusionCode1
LooPy: A Research-Friendly Mix Framework for Music Information Retrieval on Electronic Dance MusicCode1
AV-NeRF: Learning Neural Fields for Real-World Audio-Visual Scene SynthesisCode1
Differentiable Time-Frequency Scattering on GPUCode1
HiFi++: a Unified Framework for Bandwidth Extension and Speech EnhancementCode1
It's Raw! Audio Generation with State-Space ModelsCode1
Multi-Singer: Fast Multi-Singer Singing Voice Vocoder With A Large-Scale CorpusCode1
RefineGAN: Universally Generating Waveform Better than Ground Truth with Highly Accurate Pitch and Intensity ResponsesCode1
Unsupervised Source Separation By Steering Pretrained Music ModelsCode1
Taming Visually Guided Sound GenerationCode1
Neural Waveshaping SynthesisCode1
Catch-A-Waveform: Learning to Generate Audio from a Single Short ExampleCode1
Anytime Sampling for Autoregressive Models via Ordered AutoencodingCode1
Localize to Binauralize: Audio Spatialization From Visual Sound Source LocalizationCode1
Phonetic Posteriorgrams based Many-to-Many Singing Voice Conversion via Adversarial TrainingCode1
Audeo: Audio Generation for a Silent Performance VideoCode1
Perceiving Music Quality with GANsCode1
Unconditional Audio Generation with Generative Adversarial Networks and Cycle RegularizationCode1
GACELA -- A generative adversarial context encoder for long audio inpaintingCode1
Blow: a single-scale hyperconditioned flow for non-parallel raw-audio voice conversionCode1
Adversarial Audio SynthesisCode1
WaveNet: A Generative Model for Raw AudioCode1
Show:102550
← PrevPage 4 of 11Next →
All datasetsAudioCapsClassical music, 5 seconds at 12 kHzSymphony music

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1AudioGenFD_openl3185.53—Unverified
2AudioLDM2-largeFD_openl3158.04—Unverified
3Stable Audio 2.0FD_openl3110.62—Unverified
4Stable AudioFD_openl3103.66—Unverified
5ETTAFD_openl380.13—Unverified
6TangoFlux-baseFD_openl379.7—Unverified
7Stable Audio OpenFD_openl378.24—Unverified
8TangoFluxFD_openl375.1—Unverified
9ETTA-FT-AC-100kFD_openl361.79—Unverified
10DiffsoundFAD7.75—Unverified
#ModelMetricClaimedVerifiedStatus
1VAB-Encodec (Ours)Bits per byte40—Unverified
2Sparse Transformer 152M (strided)Bits per byte1.97—Unverified
#ModelMetricClaimedVerifiedStatus
1SymphonyNet Human listening average results3.5—Unverified