Champion Solution for the WSDM2023 Toloka VQA Challenge
This is an incremental improvement for visual question answering and localization tasks, specifically addressing a complex scenario in a competition setting.
The authors tackled the WSDM2023 Toloka VQA Challenge, which involves inferring and locating objects from interrogative questions, by adapting Uni-Perceiver with ViT-Adapter for cross-modal localization, achieving 77.5 and 76.347 IoU on public and private test sets to rank first.
In this report, we present our champion solution to the WSDM2023 Toloka Visual Question Answering (VQA) Challenge. Different from the common VQA and visual grounding (VG) tasks, this challenge involves a more complex scenario, i.e. inferring and locating the object implicitly specified by the given interrogative question. For this task, we leverage ViT-Adapter, a pre-training-free adapter network, to adapt multi-modal pre-trained Uni-Perceiver for better cross-modal localization. Our method ranks first on the leaderboard, achieving 77.5 and 76.347 IoU on public and private test sets, respectively. It shows that ViT-Adapter is also an effective paradigm for adapting the unified perception model to vision-language downstream tasks. Code and models will be released at https://github.com/czczup/ViT-Adapter/tree/main/wsdm2023.