VCR (Visual Commonsense Reasoning)
Introduced by Zellers et al. in From Recognition to Cognition: Visual Commonsense Reasoning
Visual Commonsense Reasoning (VCR) is a large-scale dataset for cognition-level visual understanding. Given a challenging question about an image, machines need to present two sub-tasks: answer correctly and provide a rationale justifying its answer. The VCR dataset contains over 212K (training), 26K (validation) and 25K (testing) questions, answers and rationales derived from 110K movie scenes.
Source: Visual Commonsense R-CNNBenchmarks
| Trend | Task | Dataset Variant | Best Model | Paper | Code |
|---|---|---|---|---|---|
|
VCR (Q-A) test
|
GPT4RoI
|
|||
|
VCR (QA-R) test
|
GPT4RoI
|
|||
|
VCR (Q-AR) test
|
GPT4RoI
|
|||
|
VCR (Q-A) dev
|
VL-BERTLARGE
|
|||
|
VCR (QA-R) dev
|
VL-BERTLARGE
|
|||
|
VCR (Q-AR) dev
|
VL-BERTLARGE
|
|||
|
VCR
|
OFA-X-MT
|
|||
|
VCR (QA-R) test
|
PEVL
|
|||
|
VCR (Q-A) test
|
PEVL
|
|||
|
VCR (Q-A) dev
|
PEVL
|
|||
|
VCR (QA-R) dev
|
PEVL
|
|||
|
VCR (Q-AR) dev
|
PEVL
|
|||
|
VCR (Q-AR) test
|
PEVL
|
Papers
| Paper | Code | Results | Date | Stars |
|---|
Dataset Loaders
No data loaders found. You can
submit your data loader here.
No data loaders found. You can
submit your data loader here.
