A new network improves image recognition using human visual system principles.
― 5 min read
Computer Science - Computer Vision and Pattern Recognition
RSSA new method improves low-light photography by separating brightness and noise reduction tasks.
― 7 min read
New models enhance performance in data classification tasks using concepts from sleep.
― 7 min read
A new method improves object movement analysis in challenging environments.
― 7 min read
DAPONet uses street view images to accurately identify road damage in real-time.
― 5 min read
New method improves late-stage DCE-MRI imaging for better breast cancer diagnosis.
― 6 min read
A new framework enhances motion generation for animations and virtual experiences.
― 6 min read
A new method improves hyperspectral image resolution using pre-trained RGB models.
― 5 min read
A new method improves sensor data combining accuracy and efficiency.
― 6 min read
iSeg improves image segmentation accuracy with less training data.
― 4 min read
A new model streamlines audio production by automatically eliminating breath sounds.
― 6 min read
This project tests lane-following methods for safer self-driving vehicle operation.
― 4 min read
MM-DPCNs improve video analysis efficiency by learning features without labels.
― 4 min read
MobileUNETR offers improved skin cancer detection through advanced image segmentation.
― 6 min read
A new method enhances 3D image quality using dense metric depth.
― 6 min read
A new framework enhances the representation of neural fields on triangle meshes.
― 6 min read
A method to improve action recognition with fewer labeled videos and more unlabeled data.
― 6 min read
LongLLaVA improves multi-image understanding for various applications.
― 5 min read
This article examines how combining real and synthetic images boosts face recognition accuracy and fairness.
― 5 min read
New method generates realistic 3D human models from single images using advanced video techniques.
― 5 min read
Innovative method combines machine learning and physics for solving differential equations.
― 6 min read
Introducing new metrics for assessing handwritten text generation systems.
― 6 min read
A new method improves predictions of hand movements in videos for robots and virtual reality.
― 5 min read
A new model enhances fashion item suggestions using geometry and visual data.
― 4 min read
A new method predicts BMI using handwriting styles and deep learning.
― 6 min read
Study reveals voice data's role in recognizing emotions in Spanish speakers.
― 5 min read
This study presents a model that integrates context for better facial expression recognition.
― 8 min read
New models improve road damage detection with drones, enhancing city safety.
― 5 min read
StyleTokenizer improves image generation by separating style and text instructions.
― 7 min read
This approach combines autoencoders and diffusion techniques for clearer images.
― 6 min read
Plane2Depth improves depth estimation in complex scenes, addressing challenges of low texture.
― 6 min read
This research enhances depth estimation in robots using meta-learning for better performance in varied environments.
― 5 min read
A system helps identify Korean dishes for those with dietary needs.
― 6 min read
New video generation method improves realism for self-driving car training.
― 6 min read
A new framework enhances text descriptions using images and structured data.
― 5 min read
FODA-PG enhances report generation from medical images for better diagnosis.
― 5 min read
A new method and dataset for automated cell analysis in brain research.
― 4 min read
A new approach to create synthetic images efficiently for dataset distillation.
― 8 min read
This project explores AI methods for more efficient garbage classification.
― 5 min read
This study examines the use of generative systems for managing historical photographs in Catalan archives.
― 6 min read