TextVQA

TextVQA requires models to read and reason about text in images to answer questions about them. Specifically, models need to incorporate a new modality of text present in the images and reason over it to answer TextVQA questions. TextVQA dataset contains 45,336 questions over 28,408 images from the OpenImages dataset.

Papers using TextVQA (20)

Hidden Clones: Exposing and Fixing Family Bias in Vision-Language Model Ensembles2026VOILA: Value-of-Information Guided Fidelity Selection for Cost-Aware Multimodal Question Answering2026LinMU: Multimodal Understanding Made Linear2026Text-VQA Aug: Pipelined Harnessing of Large Multimodal Models for Automated Synthesis2025Fusion to Enhance: Fusion Visual Encoder to Enhance Multimodal Language Model2025When Big Models Train Small Ones: Label-Free Model Parity Alignment for Efficient Visual Question Answering using Small VLMs2025Fast or Slow? Integrating Fast Intuition and Deliberate Thinking for Enhancing Visual Question Answering2025ASCD: Attention-Steerable Contrastive Decoding for Reducing Hallucination in MLLM2025Constructive Distortion: Improving Mllms With Attention-guided Image Warping2025Instruction-Aligned Visual Attention for Mitigating Hallucinations in Large Vision-Language Models2025Toward 3D Spatial Reasoning for Human-like Text-based Visual Question Answering2022 · 8 citesTAG: Boosting Text-VQA via Text-aware Visual Question-answer Generation2022 · 7 citesWinner Team Mia at TextVQA Challenge 2021: Vision-and-Language Representation Learning with Pre-trained Sequence-to-Sequence Model2021 · 3 citesLaTr: Layout-Aware Transformer for Scene-Text VQA2021 · 2 citesSceneGATE: Scene-Graph based co-Attention networks for TExt visual question answering2022 · 1 citesLocate Then Generate: Bridging Vision and Language with Bounding Box for Scene-Text VQA2023 · 1 citesTowards a Unified Multimodal Reasoning Framework2023Enhancing Instruction-Following Capability of Visual-Language Models by Reducing Image Redundancy2024Spatially Aware Multimodal Transformers for TextVQA2020EKTVQA: Generalized use of External Knowledge to empower Scene Text in Text-VQA2021