Blending text and images for efficient AI training.
― 6 min read
Computer Science - Computer Vision and Pattern Recognition
RSSA new method improves accuracy in recognizing Bengali handwritten text.
― 5 min read
ShapeSplat dataset enhances 3D understanding through labeled Gaussian objects.
― 7 min read
A new method to count white blood cells in muscle injuries improves healing analysis.
― 6 min read
Innovative system uses advanced models for effective road safety evaluations.
― 5 min read
Research shows people prefer larger bounding boxes in object detection.
― 6 min read
A new model improves image quality in low-dose CT scans.
― 6 min read
This new dataset tracks lake size changes every two years worldwide.
― 9 min read
This method improves 3D pose accuracy from 2D images using a transformer network.
― 5 min read
This study examines how vehicles can use AI to improve safety and communication.
― 8 min read
Simplifying object detection while increasing accuracy and efficiency.
― 5 min read
A new method improves detection of brain bleeds in trauma cases using deep learning.
― 5 min read
A new method speeds up weld seam detection with accuracy.
― 5 min read
New methods improve video segmentation accuracy and efficiency for various applications.
― 5 min read
New system enhances air-ground robot performance in dynamic environments.
― 6 min read
SpaMo improves sign language translation by focusing on visual features.
― 7 min read
MUSE improves video retrieval accuracy through multi-scale feature learning.
― 5 min read
Kalib simplifies robot calibration without physical markers, improving accuracy and efficiency.
― 5 min read
A new model predicts CST integrity from CT scans for stroke recovery.
― 5 min read
A new method improves semantic segmentation by tackling class imbalance in 3D data.
― 4 min read
NutrifyAI simplifies diet tracking through food recognition and personalized meal suggestions.
― 6 min read
A new method improves 3D scene understanding with less labeled data.
― 6 min read
New computer vision techniques aim to improve tracking of space debris and inactive satellites.
― 5 min read
A new method improves sampling and analysis of noisy 3D data.
― 5 min read
Combining local and global descriptors enhances the precision of visual localization.
― 5 min read
A new method improves predictions from fMRI data using limited samples.
― 5 min read
This system helps cameras find their position using various map techniques.
― 5 min read
IncX offers fast and clear explanations for AI object detection in real time.
― 6 min read
SynPlay enhances computer models with diverse human motion data in various scenarios.
― 5 min read
New methods safeguard images from manipulation using advanced diffusion models.
― 5 min read
A new model improves efficiency and performance in processing text and images.
― 5 min read
Study reveals strengths and weaknesses of VLMs in understanding spatial dimensions.
― 8 min read
OE-3DIS identifies and segments 3D objects without predefined names.
― 7 min read
CluMo helps models learn continuously in Visual Question Answering without forgetting past knowledge.
― 6 min read
A new method enhances medical image segmentation across different imaging techniques.
― 5 min read
A new method enhances accuracy in counting objects in generated images.
― 7 min read
A study on why deep learning models vary in performance.
― 5 min read
ControlCol enhances user control in automatic video colorization.
― 5 min read
A new method enhances assembly state recognition, reducing errors and improving efficiency.
― 7 min read
A new method enhances image quality by managing distractions in 3D graphics.
― 6 min read