# Multimodal Retrieval on VGGSound-S (test) Source: [Multimodal Retrieval on VGGSound-S (test)]() Collection index: [llms.txt]() Data updated: 2026-04-14T09:55:13.606Z Evaluation of multimodal retrieval performance (covering Video, Text, and Audio pairs) on the test split of the VGGSound-S dataset, measured using Recall@K metrics. 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 | Multimodal Retrieval | | Dataset (including any stated split) | VGGSound-S (test) | | Metric | Recall@1 (Video -\> Text) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 19 of 19 | | 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 - [Recall@1 (Video -\> Text)](): Higher is better; unit: PERCENT. - [Recall@1 (Text -\> Video)](): Higher is better; unit: PERCENT. - [Recall@1 (Video -\> Audio)](): Higher is better; unit: PERCENT. - [Recall@1 (Audio -\> Video)](): Higher is better; unit: PERCENT. - [Recall@1 (Text -\> Audio)](): Higher is better; unit: PERCENT. - [Recall@1 (Audio -\> Text)](): Higher is better; unit: PERCENT. - [Recall@1 (Average)](): Higher is better; unit: PERCENT. - [Recall@5 (Video -\> Text)](): Higher is better; unit: PERCENT. - [Recall@5 (Text -\> Video)](): Higher is better; unit: PERCENT. - [Recall@5 (Video -\> Audio)](): Higher is better; unit: PERCENT. - [Recall@5 (Audio -\> Video)](): Higher is better; unit: PERCENT. - [Recall@5 (Text -\> Audio)](): Higher is better; unit: PERCENT. - [Recall@5 (Audio -\> Text)](): Higher is better; unit: PERCENT. - [Recall@5 (Average)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Recall@1 (Video -\> Text) | Result date | Report paper | | --- | --- | --- | --- | | Full Dataset | 6.8 | Not recorded | 69ddca1a7d0172140ec79a57 | | HoPA | 6.7 | Not recorded | 69ddca1a7d0172140ec79a57 | | HoPA | 6.6 | Not recorded | 69ddca1a7d0172140ec79a57 | | HoPA | 6.6 | Not recorded | 69ddca1a7d0172140ec79a57 | | LoRS | 6.4 | Not recorded | 69ddca1a7d0172140ec79a57 | | LoRS | 6.4 | Not recorded | 69ddca1a7d0172140ec79a57 | | RepBlend | 6.4 | Not recorded | 69ddca1a7d0172140ec79a57 | | RepBlend | 6.3 | Not recorded | 69ddca1a7d0172140ec79a57 | | MTT-VL | 6.2 | Not recorded | 69ddca1a7d0172140ec79a57 | | Random | 6.1 | Not recorded | 69ddca1a7d0172140ec79a57 | | TESLA | 6.1 | Not recorded | 69ddca1a7d0172140ec79a57 | | Random | 6 | Not recorded | 69ddca1a7d0172140ec79a57 | | TESLA | 6 | Not recorded | 69ddca1a7d0172140ec79a57 | | TESLA | 5.9 | Not recorded | 69ddca1a7d0172140ec79a57 | | MTT-VL | 5.9 | Not recorded | 69ddca1a7d0172140ec79a57 | | MTT-VL | 5.8 | Not recorded | 69ddca1a7d0172140ec79a57 | | Random | 5.7 | Not recorded | 69ddca1a7d0172140ec79a57 | | RepBlend | 5.6 | Not recorded | 69ddca1a7d0172140ec79a57 | | LoRS | 5.4 | Not recorded | 69ddca1a7d0172140ec79a57 | ### Full Dataset Recall@1 (Video -\> Text): 6.8. Result ID: 69de0ef947f34ce924c5f1fd. [Reporting paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"Full"}. This record does not assert independent reproduction. ### HoPA Recall@1 (Video -\> Text): 6.7. Result ID: 69de0ef947f34ce924c5f1f6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"200 (2.5%)","Time (mins)":"50.4"}. This record does not assert independent reproduction. ### HoPA Recall@1 (Video -\> Text): 6.6. Result ID: 69de0ef947f34ce924c5f1f0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"100 (1.25%)","Time (mins)":"45.7"}. This record does not assert independent reproduction. ### HoPA Recall@1 (Video -\> Text): 6.6. Result ID: 69de0ef947f34ce924c5f1fc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"500 (6.25%)","Time (mins)":"55.1"}. This record does not assert independent reproduction. ### LoRS Recall@1 (Video -\> Text): 6.4. Result ID: 69de0ef947f34ce924c5f1f4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"200 (2.5%)","Time (mins)":"85.5"}. This record does not assert independent reproduction. ### LoRS Recall@1 (Video -\> Text): 6.4. Result ID: 69de0ef947f34ce924c5f1fa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"500 (6.25%)","Time (mins)":"124.8"}. This record does not assert independent reproduction. ### RepBlend Recall@1 (Video -\> Text): 6.4. Result ID: 69de0ef947f34ce924c5f1fb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"500 (6.25%)","Time (mins)":"125.5"}. This record does not assert independent reproduction. ### RepBlend Recall@1 (Video -\> Text): 6.3. Result ID: 69de0ef947f34ce924c5f1f5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"200 (2.5%)","Time (mins)":"87.2"}. This record does not assert independent reproduction. ### MTT-VL Recall@1 (Video -\> Text): 6.2. Result ID: 69de0ef947f34ce924c5f1ec. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"100 (1.25%)","Time (mins)":"80.9"}. This record does not assert independent reproduction. ### Random Recall@1 (Video -\> Text): 6.1. Result ID: 69de0ef947f34ce924c5f1f7. [Reporting paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"500 (6.25%)"}. This record does not assert independent reproduction. ### TESLA Recall@1 (Video -\> Text): 6.1. Result ID: 69de0ef947f34ce924c5f1f9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"500 (6.25%)","Time (mins)":"124.3"}. This record does not assert independent reproduction. ### Random Recall@1 (Video -\> Text): 6. Result ID: 69de0ef947f34ce924c5f1f1. [Reporting paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"200 (2.5%)"}. This record does not assert independent reproduction. ### TESLA Recall@1 (Video -\> Text): 6. Result ID: 69de0ef947f34ce924c5f1f3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"200 (2.5%)","Time (mins)":"85.8"}. This record does not assert independent reproduction. ### TESLA Recall@1 (Video -\> Text): 5.9. Result ID: 69de0ef947f34ce924c5f1ed. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"100 (1.25%)","Time (mins)":"81.5"}. This record does not assert independent reproduction. ### MTT-VL Recall@1 (Video -\> Text): 5.9. Result ID: 69de0ef947f34ce924c5f1f8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"500 (6.25%)","Time (mins)":"123.7"}. This record does not assert independent reproduction. ### MTT-VL Recall@1 (Video -\> Text): 5.8. Result ID: 69de0ef947f34ce924c5f1f2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"200 (2.5%)","Time (mins)":"86.3"}. This record does not assert independent reproduction. ### Random Recall@1 (Video -\> Text): 5.7. Result ID: 69de0ef947f34ce924c5f1eb. [Reporting paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"100 (1.25%)"}. This record does not assert independent reproduction. ### RepBlend Recall@1 (Video -\> Text): 5.6. Result ID: 69de0ef947f34ce924c5f1ef. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"100 (1.25%)","Time (mins)":"81.9"}. This record does not assert independent reproduction. ### LoRS Recall@1 (Video -\> Text): 5.4. Result ID: 69de0ef947f34ce924c5f1ee. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Triplets (Ratio)":"100 (1.25%)","Time (mins)":"84.9"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.