SOTAVerified

Visual Question Answering (VQA)

Visual Question Answering (VQA) is a task in computer vision that involves answering questions about an image. The goal of VQA is to teach machines to understand the content of an image and answer questions about it in natural language.

Image Source: visualqa.org

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 851–900 of 2167 papers

TitleStatusHype
Deep Equilibrium Multimodal Fusion—0
LLaVAR: Enhanced Visual Instruction Tuning for Text-Rich Image UnderstandingCode2
Answer Mining from a Pool of Images: Towards Retrieval-Based Visual Question AnsweringCode1
Pre-Training Multi-Modal Dense Retrievers for Outside-Knowledge Visual Question AnsweringCode0
Shikra: Unleashing Multimodal LLM's Referential Dialogue MagicCode2
Kosmos-2: Grounding Multimodal Large Language Models to the WorldCode1
Mitigating Hallucination in Large Multi-Modal Models via Robust Instruction TuningCode2
FunQA: Towards Surprising Video ComprehensionCode1
Visual Question Answering in Remote Sensing with Cross-Attention and Multimodal Information Bottleneck—0
StarVQA+: Co-training Space-Time Attention for Video Quality AssessmentCode0
Investigating Prompting Techniques for Zero- and Few-Shot Visual Question AnsweringCode1
Encyclopedic VQA: Visual questions about detailed properties of fine-grained categories—0
COSA: Concatenated Sample Pretrained Vision-Language Foundation ModelCode1
Improving Selective Visual Question Answering by Learning from Your PeersCode1
Scalable Neural-Probabilistic Answer Set ProgrammingCode1
Visual Question Answering (VQA) on Images with Superimposed Text—0
AVIS: Autonomous Visual Information Seeking with Large Language Model Agent—0
Weakly Supervised Visual Question Answer Generation—0
Modular Visual Question Answering via Code GenerationCode1
Knowledge Detection by Relevant Question and Image Attributes in Visual Question Answering—0
Rewarded soups: towards Pareto-optimal alignment by interpolating weights fine-tuned on diverse rewardsCode1
Q: How to Specialize Large Vision-Language Models to Data-Scarce VQA Tasks? A: Self-Train on Unlabeled Images!Code1
Multi-CLIP: Contrastive Vision-Language Pre-training for Question Answering tasks in 3D Scenes—0
DocFormerv2: Local Features for Document UnderstandingCode1
MetaVL: Transferring In-Context Learning Ability From Language Models to Vision-Language Models—0
Revisiting the Role of Language Priors in Vision-Language ModelsCode1
Evaluating the Capabilities of Multi-modal Reasoning Models with Synthetic Task Data—0
Layout and Task Aware Instruction Prompt for Zero-shot Document Image Question AnsweringCode1
LiT-4-RSVQA: Lightweight Transformer-based Visual Question Answering in Remote Sensing—0
End-to-end Knowledge Retrieval with Multi-modal QueriesCode1
Overcoming Language Bias in Remote Sensing Visual Question Answering via Adversarial Training—0
Using Visual Cropping to Enhance Fine-Detail Question Answering of BLIP-Family Models—0
Unveiling Cross Modality Bias in Visual Question Answering: A Causal View with Possible Worlds VQA—0
Generate then Select: Open-ended Visual Question Answering Guided by World Knowledge—0
VAST: A Vision-Audio-Subtitle-Text Omni-Modality Foundation Model and DatasetCode2
PaLI-X: On Scaling up a Multilingual Vision and Language ModelCode1
HaVQA: A Dataset for Visual Question Answering and Multimodal Research in Hausa LanguageCode0
CrossGET: Cross-Guided Ensemble of Tokens for Accelerating Vision-Language TransformersCode1
Modularized Zero-shot VQA with Pre-trained ModelsCode0
Study of Subjective and Objective Quality Assessment of Mobile Cloud Gaming Videos—0
Zero-shot Visual Question Answering with Language Model FeedbackCode0
Dynamic Clue Bottlenecks: Towards Interpretable-by-Design Visual Question Answering—0
NuScenes-QA: A Multi-modal Visual Question Answering Benchmark for Autonomous Driving ScenarioCode2
Measuring Faithful and Plausible Visual Grounding in VQACode0
Transferring Visual Attributes from Natural Language to Verified Image Generation—0
Image Manipulation via Multi-Hop Instructions -- A New Dataset and Weakly-Supervised Neuro-Symbolic Approach—0
DUBLIN -- Document Understanding By Language-Image Network—0
Getting ViT in Shape: Scaling Laws for Compute-Optimal Model Design—0
VLAB: Enhancing Video Language Pre-training by Feature Adapting and Blending—0
Towards Explainable In-the-Wild Video Quality Assessment: A Database and a Language-Prompted ApproachCode1
Show:102550
← PrevPage 18 of 44Next →
All datasetsGQA Test2019VQA v2 test-devVQA v2 test-stdOK-VQAMSVD-QADocVQA testMSRVTT-QAInfographicVQAGQA test-devVizWiz 2020 VQAA-OKVQACLEVR

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1humanAccuracy89.3—Unverified
2DREAM+Unicoder-VL (MSRA)Accuracy76.04—Unverified
3TRRNet (Ensemble)Accuracy74.03—Unverified
4MIL-nbgaoAccuracy73.81—Unverified
5Kakao BrainAccuracy73.33—Unverified
6Coarse-to-Fine Reasoning, Single ModelAccuracy72.14—Unverified
7270Accuracy70.23—Unverified
8NSM ensemble (updated)Accuracy67.55—Unverified
9VinVL-DPTAccuracy64.92—Unverified
10VinVL+LAccuracy64.85—Unverified
#ModelMetricClaimedVerifiedStatus
1PaLIAccuracy84.3—Unverified
2BEiT-3Accuracy84.19—Unverified
3VLMoAccuracy82.78—Unverified
4ONE-PEACEAccuracy82.6—Unverified
5mPLUG (Huge)Accuracy82.43—Unverified
6CuMo-7BAccuracy82.2—Unverified
7X2-VLM (large)Accuracy81.9—Unverified
8MMUAccuracy81.26—Unverified
9LyricsAccuracy81.2—Unverified
10InternVL-CAccuracy81.2—Unverified
#ModelMetricClaimedVerifiedStatus
1BEiT-3overall84.03—Unverified
2mPLUG-Hugeoverall83.62—Unverified
3ONE-PEACEoverall82.52—Unverified
4X2-VLM (large)overall81.8—Unverified
5VLMooverall81.3—Unverified
6SimVLMoverall80.34—Unverified
7X2-VLM (base)overall80.2—Unverified
8VASToverall80.19—Unverified
9VALORoverall78.62—Unverified
10Prompt Tuningoverall78.53—Unverified