Optimizing CNN performance on FPGAs through memory management techniques.
― 6 min read
Computer Science - Computer Vision and Pattern Recognition
RSSA new method improves action recognition in unedited videos using few examples.
― 8 min read
A new approach to depth estimation from a single image, bypassing camera limitations.
― 7 min read
COCO-ReM enhances object detection by refining COCO's original dataset annotations.
― 5 min read
Garment3DGen offers an easy method for creating 3D garments from images or text.
― 5 min read
This study focuses on automating defect analysis in silicon carbide wafers using machine learning.
― 8 min read
Annolid automates animal tracking in videos, aiding researchers in studying behavior.
― 5 min read
FlexEdit simplifies image editing by focusing on specific objects with natural results.
― 5 min read
RAP system adapts task steps from instructional videos for better planning.
― 7 min read
Examining the impact of attacks on machine learning efficiency and proposing new strategies.
― 6 min read
A new method enhances data diversity for better 3D hand models.
― 7 min read
Deep learning techniques enhance the quality of CBCT images by reducing artifacts.
― 6 min read
A new method enhances image generation based on personal input images.
― 5 min read
This study highlights offline RL methods for improving VLN agents using suboptimal data.
― 8 min read
Automated solutions for identifying brain tumors show promise in improving patient care.
― 6 min read
H-SAM improves medical image analysis with less labeled data needed.
― 4 min read
A new framework improves knowledge graph completion with diverse data types.
― 8 min read
SEMA offers a flexible approach for continual learning, adapting models to new data effectively.
― 6 min read
A new method improves visual data processing efficiency and accuracy.
― 6 min read
A new method improves camera localization accuracy while preserving user privacy.
― 5 min read
Introducing DeNetDM, a technique to reduce biases in neural networks without complex adjustments.
― 7 min read
A new method to verify deep neural networks using vision-language models.
― 7 min read
X-MIC framework enhances models for recognizing actions in first-person videos.
― 6 min read
Monitor your hydration with a smartphone camera, quick and easy.
― 5 min read
This method simplifies blending different LoRA models for high-quality image generation.
― 5 min read
New machine learning models enhance eye tracking accuracy in virtual reality environments.
― 6 min read
A new tool enhances how we interpret surface changes using remote sensing data.
― 7 min read
GANTASTIC combines GANs and diffusion models for accurate image editing.
― 5 min read
ILPO-Net offers a robust method for recognizing 3D patterns, overcoming rotation challenges.
― 6 min read
A new method improves depth measurement of see-through items using segmentation and NeRF.
― 7 min read
Diff-Mix improves image classification by creating diverse synthetic data.
― 7 min read
A new approach improves image captioning with location-aware techniques.
― 6 min read
Examining content and motion replication in video generation models.
― 8 min read
New system improves object detection using diverse data sources.
― 6 min read
A new method enhances self-supervised learning by focusing on high-quality image pairs.
― 5 min read
Introduction to GlORIE-SLAM, a new RGB-only SLAM system.
― 7 min read
Examining the emergence of features in image and audio generation models.
― 6 min read
A new method uses reinforcement learning to prune CNNs while training.
― 8 min read
A new approach improves accuracy in aligning 3D scene graphs for spatial tasks.
― 5 min read
A new model generates realistic movements in conversations, improving interaction understanding.
― 5 min read