Survey of segmentation when there are few examples

Segment new objects without forgetting old ones by replaying most informative samples with Zhu

Image augmentation using DETR transformer to select a patch with DeMix

Small object detection using sparsely-connected convolution for less compute with TinyDet

Learn new objects without forgetting old ones with transformers by using reliability with CL-DETR

Real-time segment all objects in image without training on CPU with Segment Anything

Get 3D head that is photorealistic and animatable from 1 minute of monocular video with MonoAvatar

Survey of vision-language models for vision tasks

Get 3D shape even with changing light using view-dependence normalization with VDN-NeRF

1.5x faster vision transformers by using activation sparsity with SparseViT

Get object pose in new domain without labels by automatically fine-tune on new images with TTA-COPE

Segment objects in videos using only bounding boxes along with time consistency with MaskFreeVIS

Get high-quality 3D shape with realistic texture from LiDAR smartphone by neural geometry with TMO

Real-time shape of moving 3D object and 6-DoF tracking using neural object field with BundleSDF

Complete point clouds using unsupervised learning by the Sinkhorn algorithm with UDPReg

Segment objects in videos using only 2 labeled frames with Two-shot-Video-Object-Segmentation

Detect all known and unknown objects using many sources and probability calibration with UniDetector

Get 3D shape of novel object from one image using pre-trained diffusion models with Zero-1-to-3

Vision transformer beats CNN on mobile devices for accuracy and speed with ElasticViT

Get 3D surface from Neural Radiance Field using signed surface approximation with NeRFMeshing