# Multi-sound source localization on VGGSound-Duet (test) Source: [Multi-sound source localization on VGGSound-Duet (test)]() Collection index: [llms.txt]() Data updated: 2026-04-10T00:25:32.849Z Evaluation of Multi-sound source localization performance on the test split of the VGGSound-Duet dataset, using metrics including CAP, PIAP, CIoU@0.3, and AUC. 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 | Multi-sound source localization | | Dataset (including any stated split) | VGGSound-Duet (test) | | Metric | CIoU@0.3 | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 37 of 37 | | 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 - [CIoU@0.3](): Higher is better; unit: PERCENT. - [CAP](): Higher is better; unit: PERCENT. - [PIAP](): Higher is better; unit: PERCENT. - [AUC](): Higher is better; unit: PERCENT. ## Recorded results | Method | CIoU@0.3 | Result date | Report paper | | --- | --- | --- | --- | | Generation-Analysis-Refinement | 77.6 | Not recorded | 69d732b97d0172140ec76d09 | | Generation-Analysis-Refinement | 59.5 | Not recorded | 69d732b97d0172140ec76d09 | | OA-SSL | 55.2 | Not recorded | 69d732b97d0172140ec76d09 | | Proposed Method | 46.9 | Not recorded | 68fa72d01e209131cea58af0 | | NoPrior | 46.9 | Not recorded | 69d732b97d0172140ec76d09 | | Qwen2.5-Omni | 42.6 | Not recorded | 69d732b97d0172140ec76d09 | | T-VSL | 40.1 | Not recorded | 68fa73fd1e209131cea5f120 | | MiniCPM-o | 38.6 | Not recorded | 69d732b97d0172140ec76d09 | | AVGN | 37.8 | Not recorded | 68fa73fd1e209131cea5f120 | | DSOL | 36.9 | Not recorded | 68fa73fd1e209131cea5f120 | | Alignment | 35.8 | Not recorded | 68fa73fd1e209131cea5f120 | | Mix-and-Localize | 35.6 | Not recorded | 68fa73fd1e209131cea5f120 | | FNAC | 35.1 | Not recorded | 68fa73fd1e209131cea5f120 | | AudioCLIP | 34.9 | Not recorded | 68fa73fd1e209131cea5f120 | | MarginNCE | 33.2 | Not recorded | 68fa73fd1e209131cea5f120 | | EZ-VSL | 32.4 | Not recorded | 68fa73fd1e209131cea5f120 | | LVS | 31.8 | Not recorded | 68fa73fd1e209131cea5f120 | | Wav2CLIP | 31.4 | Not recorded | 68fa73fd1e209131cea5f120 | | CoarsetoFine | 28.4 | Not recorded | 68fa73fd1e209131cea5f120 | | OTS | 26.4 | Not recorded | 68fa73fd1e209131cea5f120 | | AVGN | 26.2 | Not recorded | 68fa72d01e209131cea58af0 | | AVGN | 26.2 | Not recorded | 69d732b97d0172140ec76d09 | | DSOL | 22.3 | Not recorded | 68fa72d01e209131cea58af0 | | Mix-and-Localize | 21.1 | Not recorded | 68fa72d01e209131cea58af0 | | Mix-and-Localize | 21.1 | Not recorded | 69d732b97d0172140ec76d09 | | EZ-VSL | 20.5 | Not recorded | 68fa72d01e209131cea58af0 | | EZ-VSL | 20.5 | Not recorded | 69d732b97d0172140ec76d09 | | LVS | 17.3 | Not recorded | 68fa72d01e209131cea58af0 | | CoarseToFIne | 14.7 | Not recorded | 68fa72d01e209131cea58af0 | | CoarseToFine | 14.7 | Not recorded | 69d732b97d0172140ec76d09 | | InteractiveOmni | 14.6 | Not recorded | 69d732b97d0172140ec76d09 | | DMC | 13.8 | Not recorded | 68fa72d01e209131cea58af0 | | DMC | 13.8 | Not recorded | 69d732b97d0172140ec76d09 | | OTS | 12.2 | Not recorded | 68fa72d01e209131cea58af0 | | OTS | 12.2 | Not recorded | 69d732b97d0172140ec76d09 | | Attention10k | 11.5 | Not recorded | 68fa72d01e209131cea58af0 | | Attention 10k | 11.5 | Not recorded | 69d732b97d0172140ec76d09 | ### Generation-Analysis-Refinement CIoU@0.3: 77.6. Result ID: 69d8436347f34ce924c057a0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model category":"MLLMs","N iterations":"5"}. This record does not assert independent reproduction. ### Generation-Analysis-Refinement CIoU@0.3: 59.5. Result ID: 69d8436347f34ce924c0579f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model category":"MLLMs","N iterations":"3"}. This record does not assert independent reproduction. ### OA-SSL CIoU@0.3: 55.2. Result ID: 69d8436347f34ce924c0579b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model category":"Vision Model"}. This record does not assert independent reproduction. ### Proposed Method CIoU@0.3: 46.9. Result ID: 69a15ef035dfcabb00e69b11. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### NoPrior CIoU@0.3: 46.9. Result ID: 69d8436347f34ce924c0579a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model category":"Vision Model"}. This record does not assert independent reproduction. ### Qwen2.5-Omni CIoU@0.3: 42.6. Result ID: 69d8436347f34ce924c0579c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model category":"MLLMs"}. This record does not assert independent reproduction. ### T-VSL CIoU@0.3: 40.1. Result ID: 699327f73a6c314849141b90. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Encoders":"AudioCLIP (ResNet-50 + ESResNeXt)"}. This record does not assert independent reproduction. ### MiniCPM-o CIoU@0.3: 38.6. Result ID: 69d8436347f34ce924c0579d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model category":"MLLMs"}. This record does not assert independent reproduction. ### AVGN CIoU@0.3: 37.8. Result ID: 699327f73a6c314849141b8d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Encoders":"AudioCLIP pre-trained"}. This record does not assert independent reproduction. ### DSOL CIoU@0.3: 36.9. Result ID: 699327f73a6c314849141b8c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Encoders":"AudioCLIP pre-trained"}. This record does not assert independent reproduction. ### Alignment CIoU@0.3: 35.8. Result ID: 699327f73a6c314849141b8a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Encoders":"AudioCLIP pre-trained"}. This record does not assert independent reproduction. ### Mix-and-Localize CIoU@0.3: 35.6. Result ID: 699327f73a6c314849141b8b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Encoders":"AudioCLIP pre-trained"}. This record does not assert independent reproduction. ### FNAC CIoU@0.3: 35.1. Result ID: 699327f73a6c314849141b89. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Encoders":"AudioCLIP pre-trained"}. This record does not assert independent reproduction. ### AudioCLIP CIoU@0.3: 34.9. Result ID: 699327f73a6c314849141b8e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Encoders":"AudioCLIP pre-trained"}. This record does not assert independent reproduction. ### MarginNCE CIoU@0.3: 33.2. Result ID: 699327f73a6c314849141b88. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Encoders":"AudioCLIP pre-trained"}. This record does not assert independent reproduction. ### EZ-VSL CIoU@0.3: 32.4. Result ID: 699327f73a6c314849141b87. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Encoders":"AudioCLIP pre-trained"}. This record does not assert independent reproduction. ### LVS CIoU@0.3: 31.8. Result ID: 699327f73a6c314849141b86. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Encoders":"AudioCLIP pre-trained"}. This record does not assert independent reproduction. ### Wav2CLIP CIoU@0.3: 31.4. Result ID: 699327f73a6c314849141b8f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Encoders":"AudioCLIP pre-trained"}. This record does not assert independent reproduction. ### CoarsetoFine CIoU@0.3: 28.4. Result ID: 699327f73a6c314849141b85. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Encoders":"AudioCLIP pre-trained"}. This record does not assert independent reproduction. ### OTS CIoU@0.3: 26.4. Result ID: 699327f73a6c314849141b84. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Encoders":"AudioCLIP pre-trained"}. This record does not assert independent reproduction. ### AVGN CIoU@0.3: 26.2. Result ID: 69a15ef035dfcabb00e69b10. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### AVGN CIoU@0.3: 26.2. Result ID: 69d8436347f34ce924c05799. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model category":"Vision Model"}. This record does not assert independent reproduction. ### DSOL CIoU@0.3: 22.3. Result ID: 69a15ef035dfcabb00e69b0c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Mix-and-Localize CIoU@0.3: 21.1. Result ID: 69a15ef035dfcabb00e69b0f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Mix-and-Localize CIoU@0.3: 21.1. Result ID: 69d8436347f34ce924c05798. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model category":"Vision Model"}. This record does not assert independent reproduction. ### EZ-VSL CIoU@0.3: 20.5. Result ID: 69a15ef035dfcabb00e69b0e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### EZ-VSL CIoU@0.3: 20.5. Result ID: 69d8436347f34ce924c05797. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model category":"Vision Model"}. This record does not assert independent reproduction. ### LVS CIoU@0.3: 17.3. Result ID: 69a15ef035dfcabb00e69b0d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### CoarseToFIne CIoU@0.3: 14.7. Result ID: 69a15ef035dfcabb00e69b0b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### CoarseToFine CIoU@0.3: 14.7. Result ID: 69d8436347f34ce924c05796. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model category":"Vision Model"}. This record does not assert independent reproduction. ### InteractiveOmni CIoU@0.3: 14.6. Result ID: 69d8436347f34ce924c0579e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model category":"MLLMs"}. This record does not assert independent reproduction. ### DMC CIoU@0.3: 13.8. Result ID: 69a15ef035dfcabb00e69b0a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### DMC CIoU@0.3: 13.8. Result ID: 69d8436347f34ce924c05795. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model category":"Vision Model"}. This record does not assert independent reproduction. ### OTS CIoU@0.3: 12.2. Result ID: 69a15ef035dfcabb00e69b09. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### OTS CIoU@0.3: 12.2. Result ID: 69d8436347f34ce924c05794. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model category":"Vision Model"}. This record does not assert independent reproduction. ### Attention10k CIoU@0.3: 11.5. Result ID: 69a15ef035dfcabb00e69b08. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Attention 10k CIoU@0.3: 11.5. Result ID: 69d8436347f34ce924c05793. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model category":"Vision Model"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.