A study evaluates various fairness improvement techniques for deep learning models in image classification.
― 8 min read
Computer Science - Computer Vision and Pattern Recognition
RSSNew model improves tracking and recognition of text in video content.
― 4 min read
A new framework enhances model performance on unseen data using targeted changes.
― 6 min read
New skin tone metrics aim to improve fairness and privacy in face recognition systems.
― 7 min read
Introducing Neat, an efficient method for classifying known and unknown data categories.
― 7 min read
A new model enhances how self-driving cars predict pedestrian movements in crowded areas.
― 5 min read
New methods generate realistic 3D avatars from a single image.
― 6 min read
A new method to improve jump cuts in video editing.
― 4 min read
A new approach to pruning datasets improves model performance while reducing data size.
― 7 min read
A new dataset for improving e-commerce image and text recognition.
― 7 min read
DiffSketch allows artists to create stylized sketches from just one example.
― 5 min read
MatSAM enhances microstructure extraction from material images for improved analysis and research.
― 7 min read
This research introduces a new method for reliable object detection in complex environments.
― 9 min read
A novel approach using diffusion models for better video scene representation.
― 6 min read
A new AI method improves mammogram analysis without needing extensive labeled data.
― 6 min read
This method simplifies adding objects to 3D environments using text and 2D boxes.
― 6 min read
Researchers use satellite images and AI to create annual forest height maps.
― 6 min read
Examining how different models for images and text can work together effectively.
― 6 min read
A new video analysis system aims to detect lameness in cows efficiently.
― 6 min read
This study enhances the analysis of CT scans using AI for better COVID-19 detection.
― 5 min read
TRIPS enhances efficiency in vision-language tasks by selecting relevant image patches.
― 7 min read
New dataset highlights the stereotype issue in image generation models.
― 6 min read
Introducing Diff-NCA and FourierDiff-NCA for efficient image generation.
― 6 min read
Learn how Variational Diffusion Models generate high-quality images and data.
― 6 min read
Research highlights the use of self-supervised pretraining in GIE image analysis.
― 6 min read
This study examines issues in models responding to visual questions.
― 6 min read
A new method improving 3D image quality using wavelet integration with Triplane.
― 6 min read
New method improves video captioning using image-language models.
― 6 min read
A new method of dubbing enhances lip syncing for better viewer experience.
― 6 min read
Research reveals hidden manipulation risks in Activation Maximization methods for DNNs.
― 8 min read
A new method enhances shadow generation for more realistic 3D character models.
― 7 min read
A fresh method improves video quality while reducing resource use.
― 5 min read
Learn how ConKeD improves retinal image alignment for better disease tracking.
― 6 min read
Study enhances coffee leaf disease classification using tech advancements and data techniques.
― 9 min read
A new method to improve efficiency in video editing through focused resource allocation.
― 8 min read
A new approach for detecting anomalies in surveillance videos with clear explanations.
― 6 min read
New methods improve image editing speed and quality using smaller models.
― 5 min read
New techniques improve anomaly detection in visual inspections using machine learning.
― 5 min read
A new method for efficient segmentation of large 3D environments.
― 6 min read
This article discusses an embedded system for effective planogram compliance in stores.
― 6 min read