Federated learning has become a popular method to learn from decentraliz...
Real-scanned point clouds are often incomplete due to viewpoint, occlusi...
Image-based 3D detection is an indispensable component of the perception...
Image-to-image (I2I) translation is a challenging topic in computer visi...
Video highlight detection is a crucial yet challenging problem that aims...
Knowledge distillation (KD) has shown very promising capabilities in
tra...
Given labeled data in a source domain, unsupervised domain adaptation ha...
The recently proposed MaskFormer <cit.> gives a refreshed
perspective on...
3D human shape and pose estimation is the essential task for human motio...
Group activity recognition is a crucial yet challenging problem, whose c...
Unsupervised representation learning has achieved outstanding performanc...
As a crucial task of autonomous driving, 3D object detection has made gr...
Deep learning-based 3D object detection has achieved unprecedented succe...
Deep learning-based multi-source unsupervised domain adaptation (MUDA) h...
Unsupervised domain adaptation has been widely adopted to generalize mod...
Most 3D shape completion approaches rely heavily on partial-complete sha...
Real-scanned point clouds are often incomplete due to viewpoint, occlusi...
Estimating 3D bounding boxes from monocular images is an essential compo...
Person Re-Identification (Re-ID) is of great importance to the many vide...
Federated learning is a privacy-preserving machine learning technique th...
Few-shot segmentation focuses on the generalization of models to segment...
We present an interesting and challenging dataset that features a large
...
Deep classifiers have achieved great success in visual recognition. Howe...
Understanding crowd motion dynamics is critical to real-world applicatio...
In this paper, we propose a general and efficient pre-training paradigm,...
Although person re-identification (ReID) has achieved significant improv...
Connectionist Temporal Classification (CTC) and attention mechanism are ...
Neural Architecture Search (NAS) achieves significant progress in many
c...
The basic principles in designing convolutional neural network (CNN)
str...
Recent works have been applying self-attention to various fields in comp...
Person re-identification (reID) is an important task that requires to
re...
Dense depth cues are important and have wide applications in various com...
Monocular depth estimation aims at estimating a pixelwise depth map for ...
Person re-identification aims to robustly measure similarities between p...
Person re-identification aims at finding a person of interest in an imag...
The person re-identification task requires to robustly estimate visual
s...
Pedestrian analysis plays a vital role in intelligent video surveillance...
Vehicle re-identification is an important problem and has many applicati...
This paper introduces the system we developed for the Youtube-8M Video
U...