TransformMix improves data augmentation by preserving essential features in mixed images.
― 6 min read
Computer Science - Computer Vision and Pattern Recognition
RSSNew framework improves human mesh recovery from multiple images.
― 7 min read
Researchers propose training models on altered images to minimize copyright issues.
― 5 min read
A new approach combining weight and token pruning for efficient image processing.
― 6 min read
ACDG-VTON improves online clothes fitting with accurate image generation.
― 5 min read
This article presents a new method for improving visual grounding by merging data and model knowledge.
― 8 min read
DepthFM improves speed and accuracy in depth estimation using flow matching techniques.
― 6 min read
An overview of how machines learn from video data and the challenges they face.
― 7 min read
New methods enhance music score reading and interpretation.
― 6 min read
Federated learning enhances cooperation between autonomous vehicles in challenging weather.
― 5 min read
DVMNet enhances object pose estimation for unseen objects with new techniques.
― 6 min read
A new framework improves pose estimation for diverse objects using meta-point learning.
― 7 min read
New framework improves object detection in urban environments using open-vocabulary learning.
― 6 min read
This method enhances StyleGAN's efficiency while preserving image variety.
― 6 min read
New model enhances CT image quality using low-resolution images.
― 5 min read
Improving motion prediction techniques for safer autonomous driving systems.
― 6 min read
A new method enhances image quality when capturing scenes in motion.
― 5 min read
A new method improves image quality affected by speckle noise using machine learning.
― 7 min read
This article discusses the benefits of RoPE for Vision Transformers in various tasks.
― 4 min read
UP-CrackNet offers a new approach to detecting road cracks with less human involvement.
― 5 min read
New technology allows users to create 3D models from text descriptions effortlessly.
― 5 min read
A new method enhances panoptic segmentation by integrating depth data.
― 6 min read
Introducing a new method for more accurate scene graph generation.
― 6 min read
New method improves clarity of underwater images using color and depth information.
― 6 min read
A new method enhances fashion image editing using multiple data sources.
― 5 min read
Active learning reduces labeling costs while improving model performance.
― 6 min read
A new approach for improving machine learning systems in image recognition tasks.
― 6 min read
This research highlights sparse coding's role in safeguarding neural network privacy.
― 6 min read
Research reveals weaknesses in 3D models' language understanding and proposes solutions.
― 8 min read
New model improves accuracy in cloud property retrieval from satellite data.
― 7 min read
MVSplat revolutionizes 3D reconstruction using sparse images with efficient techniques.
― 5 min read
LiFT improves Vision Transformers for detailed image tasks efficiently.
― 6 min read
New method enhances 3D segmentation using inconsistent 2D labels for better reconstruction.
― 7 min read
A new method enhances object detection by combining text and visual inputs.
― 5 min read
A new method for identifying anomalies in medical imaging through pseudo-healthy reconstruction.
― 6 min read
B-LoRA separates style and content in images for better artistic expression.
― 7 min read
DINO Tracker improves video point tracking, handling occlusions effectively.
― 7 min read
New techniques improve shape analysis of anatomical structures for better medical assessments.
― 6 min read
ObjectDR generates paired data to improve 3D shape reconstruction from 2D images.
― 5 min read
New method enhances identification of unusual events in video footage.
― 8 min read