ICCV 2019 accepted papers on Vision-language and multimodal
58 of the 1,075 papers accepted at ICCV 2019 (5.4%) are on Vision-language and multimodal, compared with 6.8% at ICCV 2017. They fall into 6 topics.
See ICCV 2019 in the interactive map
Submitting to ICCV? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Human and embodied modeling · 22 papers
- Attention on Attention for Image Captioning (1,117 citations)
- What Is Wrong With Scene Text Recognition Model Comparisons? Dataset and Model Analysis (666 citations)
- Visual Semantic Reasoning for Image-Text Matching (644 citations)
- A Fast and Accurate One-Stage Approach to Visual Grounding (441 citations)
- Entangled Transformer for Image Captioning (412 citations)
- Learning to Assemble Neural Module Tree Networks for Visual Grounding (312 citations)
- Robust Change Captioning (223 citations)
- Symmetry-Constrained Rectification Network for Scene Text Recognition (159 citations)
- Reflective Decoding Network for Image Captioning (115 citations)
- Align, Attend and Locate: Chest X-Ray Diagnosis via Contrast Induced Attention Network With Limited Supervision (105 citations)
- Learning to Collocate Neural Modules for Image Captioning (102 citations)
- TextPlace: Visual Place Recognition and Topological Localization Through Reading Scene Texts (61 citations)
and 10 more in the interactive map.
Cross-modal learning · 17 papers
- CAMP: Cross-Modal Adaptive Message Passing for Text-Image Retrieval (380 citations)
- Deep Joint-Semantics Reconstructing Hashing for Large-Scale Unsupervised Cross-Modal Retrieval (371 citations)
- EPIC-Fusion: Audio-Visual Temporal Binding for Egocentric Action Recognition (331 citations)
- Weakly Aligned Cross-Modal Learning for Multispectral Pedestrian Detection (306 citations)
- Adversarial Representation Learning for Text-to-Image Matching (259 citations)
- Dual Attention Matching for Audio-Visual Event Localization (221 citations)
- The Sound of Motions (217 citations)
- Fine-Grained Action Retrieval Through Multiple Parts-of-Speech Embeddings (156 citations)
- Saliency-Guided Attention Network for Image-Sentence Matching (121 citations)
- MMAct: A Large-Scale Dataset for Cross Modal Human Action Understanding (103 citations)
- Watch, Listen and Tell: Multi-Modal Weakly Supervised Dense Event Captioning (99 citations)
- Recursive Visual Sound Separation Using Minus-Plus Net (81 citations)
and 5 more in the interactive map.
Visual question answering · 12 papers
- Relation-Aware Graph Attention Network for Visual Question Answering (414 citations)
- Scene Text Visual Question Answering (358 citations)
- Taking a HINT: Leveraging Explanations to Make Vision and Language Models More Grounded (263 citations)
- Multi-Modality Latent Interaction Network for Visual Question Answering (87 citations)
- VrR-VG: Refocusing Visually-Relevant Relationships (83 citations)
- U-CAM: Visual Explanation Using Uncertainty Based Class Activation Maps (80 citations)
- Compact Trilinear Interaction for Visual Question Answering (78 citations)
- Making History Matter: History-Advantage Sequence Training for Visual Dialog (73 citations)
- Why Does a Visual Question Have Different Answers? (61 citations)
- From Strings to Things: Knowledge-Enabled VQA Model That Can Read and Reason (54 citations)
- SegEQA: Video Segmentation Based Visual Attention for Embodied Question Answering (27 citations)
- AttentionRNN: A Structured Spatial Attention Mechanism (9 citations)
CLIP and zero-shot · 4 papers
- Collect and Select: Semantic Alignment Metric Learning for Few-Shot Learning (170 citations)
- Towards Unsupervised Image Captioning With Shared Multimodal Embeddings (108 citations)
- Person Search by Text Attribute Query As Zero-Shot Learning (45 citations)
- ViCo: Word Embeddings From Visual Co-Occurrences (21 citations)
