Restore aerial images having low visibility using a lightweight algorithm with LVRNet

Remove haze in an image with detail-enhanced convolution and content-guided attention with DEA-Net

Complete point clouds with geometry-aware transformers with AdaPoinTr

Train object segmentation with only boxes by using transformers as mask auto-labeler with MAL

Survey of image noise removal using deep learning

Do many types of video segmentation with one model without retraining with TarViS

Detect known and unknown object by first localize and then identify with CAT

Improve crowd counting by varying the weight with Mask Focal Loss

Enhance blurry images with unknown turbulence without paired training data with TurbuGAN

Survey of human segmentation into parts using deep learning

Survey of 3D point cloud completion using deep learning

Estimate object size in the wild without camera calibration or handcrafted features with PMODE

Survey of self-supervised learning for videos

Better object orientation using self-supervision by learning textures with TexPose

Survey of face recognition

Improve object detection using object co-occurrence statistics with GPR

Better object segmentation using foreground learned with unlabeled images with C2AM

Survey of low-light image and video enhancement

Survey of graph neural networks in computer vision with architectures, datasets, common approaches

Survey of Neural Radiance Field in 3D Vision