# Visual Question Answering on abstract scenes (test) Source: [Visual Question Answering on abstract scenes (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:54:28.032Z Evaluation of model performance on the test split of the Abstract Scenes dataset for Visual Question Answering. 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 | Visual Question Answering | | Dataset (including any stated split) | abstract scenes (test) | | Metric | Overall Accuracy (MC) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 6 of 6 | | 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 - [Overall Accuracy (MC)](): Higher is better; unit: PERCENT. - [Yes/No Accuracy (MC)](): Higher is better; unit: PERCENT. - [Other Accuracy (MC)](): Higher is better; unit: PERCENT. - [Number Accuracy (MC)](): Higher is better; unit: PERCENT. - [Overall Accuracy (OE)](): Higher is better; unit: PERCENT. - [Yes/No Accuracy (OE)](): Higher is better; unit: PERCENT. - [Other Accuracy (OE)](): Higher is better; unit: PERCENT. - [Number Accuracy (OE)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Overall Accuracy (MC) | Result date | Report paper | | --- | --- | --- | --- | | Graph VQA | 74.37 | Not recorded | 68fa633a1e209131ce9fda20 | | U. Tokyo MIL | 71.18 | Not recorded | 68fa633a1e209131ce9fda20 | | LSTM with global image features | 69.21 | Not recorded | 68fa633a1e209131ce9fda20 | | Multimodal residual learning | 67.99 | Not recorded | 68fa633a1e209131ce9fda20 | | LSTM blind | 61.41 | Not recorded | 68fa633a1e209131ce9fda20 | | Zhang et al. | 35.25 | Not recorded | 68fa633a1e209131ce9fda20 | ### Graph VQA Overall Accuracy (MC): 74.37. Result ID: 698becf9ecace47ef55a3f8a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"model\_version":"full model"}. This record does not assert independent reproduction. ### U. Tokyo MIL Overall Accuracy (MC): 71.18. Result ID: 698becf9ecace47ef55a3f89. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"ensemble":true}. This record does not assert independent reproduction. ### LSTM with global image features Overall Accuracy (MC): 69.21. Result ID: 698becf9ecace47ef55a3f86. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"features":"global image features"}. This record does not assert independent reproduction. ### Multimodal residual learning Overall Accuracy (MC): 67.99. Result ID: 698becf9ecace47ef55a3f88. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### LSTM blind Overall Accuracy (MC): 61.41. Result ID: 698becf9ecace47ef55a3f85. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"type":"blind (language only)"}. This record does not assert independent reproduction. ### Zhang et al. Overall Accuracy (MC): 35.25. Result ID: 698becf9ecace47ef55a3f87. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"evaluation scope":"yes/no only"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.