SportsHHI focuses on human interactions in basketball and volleyball videos for improved analysis.
― 5 min read
Computer Science - Computer Vision and Pattern Recognition
RSSContext enhances video summaries, making them more informative and engaging.
― 5 min read
DTC123 improves 3D model generation from single images using teaching models.
― 6 min read
A novel approach to generate detailed images of people in complex scenes.
― 6 min read
A new lightweight model improves target recognition in synthetic aperture radar images.
― 5 min read
New method simplifies 3D scene editing using text-based prompts and depth information.
― 6 min read
Innovative methods improve image quality while reducing data usage.
― 9 min read
A new approach to merging sensor data enhances object detection and mapping.
― 5 min read
A new method uses motion to enhance video scene understanding.
― 6 min read
Using historical LiDAR data to enhance camera-based 3D detection in autonomous vehicles.
― 7 min read
A new method for tracking objects in videos without costly labeled data.
― 8 min read
New dataset and model enhance understanding of facial emotions and expressions.
― 7 min read
Research focuses on enhancing mapping methods for efficient location identification.
― 5 min read
MagicTime transforms written descriptions into dynamic time-lapse videos with improved realism.
― 6 min read
Introducing Dynamic Distinction Learning for improved anomaly detection in surveillance videos.
― 9 min read
This approach improves segmentation of medical images without extensive labeled data.
― 5 min read
Discover how SDF2Net improves PolSAR image analysis and classification accuracy.
― 5 min read
CodecNeRF improves 3D representations with fast encoding and high-quality images.
― 9 min read
A new dataset combining images and LiDAR data for advanced 3D reconstruction.
― 7 min read
Discover how human-object contact can improve 3D modeling from images.
― 4 min read
MemFlow offers real-time optical flow estimation using a memory module for enhanced accuracy.
― 6 min read
EasyTrack and EasyTrack++ improve object tracking in 3D environments using point clouds.
― 5 min read
A new model improves liver lesion classification in 3D imaging.
― 6 min read
A new approach to medical image analysis using federated learning and weak supervision.
― 5 min read
This research reveals task vectors that enhance visual model performance without extra examples.
― 9 min read
MoMA simplifies personalized image creation using just one photo and a text description.
― 5 min read
NAF-DPM enhances document clarity for better OCR accuracy.
― 7 min read
YaART creates quality images from text quickly and efficiently.
― 5 min read
Introducing a method for task-agnostic pruning of complex models.
― 7 min read
A new tool for making images safer and more responsible.
― 7 min read
Data pruning improves model efficiency while addressing potential bias issues.
― 7 min read
New methods enhance accuracy of hand position detection during interactions.
― 5 min read
A look at the benefits of Semantic-aware Classifier-Free Guidance in image creation.
― 6 min read
This article discusses a system for detecting game states in Foosball.
― 6 min read
WebXR transforms how we engage with immersive digital environments.
― 8 min read
MindSet: Vision helps researchers evaluate deep neural networks against human visual perception.
― 6 min read
This method improves hyperspectral band selection by using LiDAR data for analysis.
― 5 min read
New methods allow stunning art style personalization in image generation.
― 6 min read
CustomListener creates realistic avatars that respond to conversations dynamically.
― 6 min read
A new loss function improves accuracy in semantic segmentation tasks.
― 6 min read