# Impact sound synthesis on Greatest Hits (test) Source: [Impact sound synthesis on Greatest Hits (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:47:44.245Z Evaluation of models performing Impact sound synthesis on the test split of the Greatest Hits dataset, assessed using objective metrics like FID, KID, and Recognition Accuracy. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Impact sound synthesis | | Dataset (including any stated split) | Greatest Hits (test) | | Metric | KL Div. | | Unit | SCALAR | | Direction | Lower is better | | Results loaded | 13 of 13 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [KL Div.](): Lower is better; unit: SCALAR. - [FID](): Lower is better; unit: SCALAR. - [KID (mean)](): Lower is better; unit: SCALAR. - [KID (std)](): Lower is better; unit: SCALAR. - [Recog. Acc](): Higher is better; unit: PERCENT. - [Spec. MSE](): Lower is better; unit: SCALAR. - [Matching](): Higher is better; unit: SCALAR. - [Quality Score](): Higher is better; unit: SCALAR. ## Recorded results | Method | KL Div. | Result date | Report paper | | --- | --- | --- | --- | | Video + Physics Diffusion | 2.04 | Not recorded | 68fa6ffb1e209131cea484bc | | Physics-Driven Diffusion Model | 2.04 | Not recorded | 68fa6ffb1e209131cea484bc | | Video + Class label Diffusion | 2.38 | Not recorded | 68fa6ffb1e209131cea484bc | | Video + Spec Diffusion | 2.55 | Not recorded | 68fa6ffb1e209131cea484bc | | Video Diffusion | 2.77 | Not recorded | 68fa6ffb1e209131cea484bc | | Video + MFCC Diffusion | 2.84 | Not recorded | 68fa6ffb1e209131cea484bc | | Transformer-based | 3.13 | Not recorded | 68fa6ffb1e209131cea484bc | | ConvNet-based | 4.65 | Not recorded | 68fa6ffb1e209131cea484bc | | NN via Predicted Sound Features | 7.39 | Not recorded | 68fa6ffb1e209131cea484bc | | NN via Visual Features | 10.6 | Not recorded | 68fa6ffb1e209131cea484bc | ### Video + Physics Diffusion KL Div.: 2.04. Result ID: 699104383a6c314849050b12. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Conditioning":"Physics"}. This record does not assert independent reproduction. ### Physics-Driven Diffusion Model KL Div.: 2.04. Result ID: 699104383a6c314849050b15. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Conditioning":"Physics Priors","Model Type":"Diffusion"}. This record does not assert independent reproduction. ### Video + Class label Diffusion KL Div.: 2.38. Result ID: 699104383a6c314849050b0f. [Reporting paper]() Recorded evaluation configuration: {"Conditioning":"Class label"}. This record does not assert independent reproduction. ### Video + Spec Diffusion KL Div.: 2.55. Result ID: 699104383a6c314849050b11. [Reporting paper]() Recorded evaluation configuration: {"Conditioning":"Spectrogram"}. This record does not assert independent reproduction. ### Video Diffusion KL Div.: 2.77. Result ID: 699104383a6c314849050b0e. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Video + MFCC Diffusion KL Div.: 2.84. Result ID: 699104383a6c314849050b10. [Reporting paper]() Recorded evaluation configuration: {"Conditioning":"MFCC"}. This record does not assert independent reproduction. ### Transformer-based KL Div.: 3.13. Result ID: 699104383a6c314849050b0d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### ConvNet-based KL Div.: 4.65. Result ID: 699104383a6c314849050b0c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### NN via Predicted Sound Features KL Div.: 7.39. Result ID: 699104383a6c314849050b14. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Architecture":"LSTM","Predicted Features":"Cochleagrams"}. This record does not assert independent reproduction. ### NN via Visual Features KL Div.: 10.6. Result ID: 699104383a6c314849050b13. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50","Physics Priors":false,"Diffusion Model":false}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.