ACM Multimedia 2020

497 papers
ACM Multimedia 2020 featured 812 papers focusing on spatio-temporal scene graphs for dynamic scene and video understanding. Research directions included adversarial alignment for domain invariant feature learning, neighborhood consensus for label refinement, and the use of single stream transformers for joint feature interaction. The program addressed challenges in multimodal translation, human pose estimation, and vision robustness under adverse environmental conditions. Prolific contributors included researchers such as Zheng-Jun Zha, Qingming Huang, Wu Liu, and Meng Wang. Specific studies explored deepfake detection via audio-visual dissonance, compressed video action recognition, and infrared-visible person re-identification. Other work involved traffic-aware vehicle tracking, automated cooking workflow construction, and privacy-preserving deep learning. The edition also highlighted systems for augmented reality remote interaction and panoptic image annotation. These contributions reflect a broader trend toward integrating distributed sensor networks and resolving linguistic underspecification in multimodal datasets.

Top Research Directions

Research directions most represented by ACM Multimedia 2020 papers.

Adversarial Training for Domain Invariant Representations 4
Generalizing from Synthetic Data to the Real World 4
Mitigating Error Propagation in Pseudo Labeling 3
Preserving Background and Structure in Local Image Editing 2
Stochastic Generation for Audio Driven Motion Synthesis 2
Resolving Kinematic Ambiguity in Skeleton Based Action Recognition 2
Composite Loss Functions for Perceptual Image Restoration 2
Relational Scene Graphs for Visual Reasoning 2
Transparent Reasoning for Personalized Recommendation 2
Upstream Detection Bottlenecks in Structured Visual Reasoning 2
Asymmetric Spatiotemporal Decoupling for Video Models 2
Bidirectional Attention for Mutual Cross-Modal Refinement 2
Cross-Modal Attention Fusion for Heterogeneous Data Integration 2
Over-Smoothing in Graph Neural Networks 2
The Cross Modal Representation Gap 2

Top Authors

AuthorPapers
Zheng-Jun Zha15
Qingming Huang12
Tao Mei10
Wu Liu10
Fei Wu8
Shuhui Wang8
Meng Wang7
Tat-Seng Chua7
Jinhui Tang6
Liang Li6
Nicu Sebe6
Shuicheng YAN6
Yongdong Zhang6
Zhou Zhao6
Anan Liu5
Fei Wu5
Heng Tao Shen5
Jingjing Chen5
Liqiang Nie5
Qianqian Xu5

Impactful Papers

Papers from ACM Multimedia 2020 that shaped research clusters.

Foundational

Influential

Papers

Title
Self-Mimic Learning for Small-scale Pedestrian Detection
Self-Play Reinforcement Learning for Fast Image Retargeting
Self-supervised Dance Video Synthesis Conditioned on Music
Self-supervised Video Representation Learning Using Inter-intra Contrastive Framework
Semantic Consistency Guided Instance Feature Alignment for 2D Image-Based 3D Shape Retrieval
Semantic Image Analogy with a Conditional Single-Image GAN
Semi-supervised Multi-modal Emotion Recognition with Cross-Modal Distribution Matching
Semi-supervised Online Multi-Task Metric Learning for Visual Recognition and Retrieval
Sequential Attention GAN for Interactive Image Editing
ShapeCaptioner: Generative Caption Network for 3D Shapes by Learning a Mapping from Parts Detected in Multiple Views to Sentences
Sharp Multiple Instance Learning for DeepFake Video Detection
Siamese Attentive Graph Tracking
SimSwap: An Efficient Framework For High Fidelity Face Swapping
Simultaneous Semantic Alignment Network for Heterogeneous Domain Adaptation
Single Image De-noising via Staged Memory Network
Single Image Deraining via Scale-space Invariant Attention Neural Network
SketchMan: Learning to Create Professional Sketches
Skin Textural Generation via Blue-noise Gabor Filtering based Generative Adversarial Network
SmartShots: Enabling Automatic Generation of Videos with Data Visualizations Embedded
SOMHunter: Lightweight Video Search System with SOM-Guided Relevance Feedback
Space-Time Video Super-Resolution Using Temporal Profiles
SpatialGAN: Progressive Image Generation Based on Spatial Recursive Adversarial Expansion
Spatial-Temporal Knowledge Integration: Robust Self-Supervised Facial Landmark Tracking
Spatio-Temporal Inception Graph Convolutional Networks for Skeleton-Based Action Recognition
SRHEN: Stepwise-Refining Homography Estimation Network via Parsing Geometric Correspondences in Deep Latent Space
SST-EmotionNet: Spatial-Spectral-Temporal based Attention 3D Dense Network for EEG Emotion Recognition
Stable Video Style Transfer Based on Partial Convolution with Depth-Aware Supervision
Story Semantic Relationships from Multimodal Cognitions
Stronger, Faster and More Explainable: A Graph Convolutional Baseline for Skeleton-based Action Recognition
STRONG: Spatio-Temporal Reinforcement Learning for Cross-Modal Video Moment Localization
Structural Semantic Adversarial Active Learning for Image Captioning
SUMAC 2020: The 2nd Workshop on Structuring and Understanding of Multimedia heritAge Contents
Summary of MuSe 2020: Multimodal Sentiment Analysis, Emotion-target Engagement and Trustworthiness Detection in Real-life Media
Supervised Hierarchical Deep Hashing for Cross-Modal Retrieval
Surface Reconstruction with Unconnected Normal Maps: An Efficient Mesh-based Approach
Surpassing Real-World Source Training Data: Random 3D Characters for Generalizable Person Re-Identification
Tactile Sketch Saliency
Talking Face Generation with Expression-Tailored Generative Adversarial Network
Task Decoupled Knowledge Distillation For Lightweight Face Detectors
Task-distribution-aware Meta-learning for Cold-start CTR Prediction
Temporal Denoising Mask Synthesis Network for Learning Blind Video Temporal Consistency
Text-Embedded Bilinear Model for Fine-Grained Visual Recognition
Text-Guided Image Inpainting
Text-Guided Neural Image Inpainting
TextRay: Contour-based Geometric Modeling for Arbitrary-shaped Scene Text Detection
Text-to-Image Synthesis via Aesthetic Layout
Textual Dependency Embedding for Person Search by Language
Texture Semantically Aligned with Visibility-aware for Partial Person Re-identification
The World has Changed - The World Needs to Change. What Multimedia has to Offer for Our Common Digital Future
ThumbNet: One Thumbnail Image Contains All You Need for Recognition

« prev Page 9/10 next »