In LiDAR-based 3D detection, history point clouds contain rich temporal
...
The training process of ReLU neural networks often exhibits complicated
...
In the field of 3D object detection for autonomous driving, the sensor
p...
The performance of point cloud 3D object detection hinges on effectively...
RGB-D saliency detection aims to fuse multi-modal cues to accurately loc...
This letter proposes an analytical framework to evaluate the coverage
pe...
Deep learning models have shown their vulnerability when dealing with
ad...
In this paper, we show that structures similar to self-attention are nat...
In this work, we explore the maximum-margin bias of quasi-homogeneous ne...
Due to the difficulty in collecting paired real-world training data, ima...
A general framework with a series of different methods is proposed to im...
High-speed, high-resolution stereoscopic (H2-Stereo) video allows us to
...
Transformer trackers have achieved impressive advancements recently, whe...
Recent RGB-D semantic segmentation has motivated research interest thank...
Generalization error bounds for deep neural networks trained by stochast...
The convergence of GD and SGD when training mildly parameterized neural
...
Real-time and high-performance 3D object detection is of critical import...
Local quadratic approximation has been extensively used to study the
opt...
Various facial manipulation techniques have drawn serious public concern...
In this paper, we study the problem of finding mixed Nash equilibrium fo...
Variational Autoencoders (VAEs) have recently been highly successful at
...
Graph neural networks (GNN) have shown great success in learning from
gr...
The training dynamics of two-layer neural networks with batch normalizat...
Recent RGBD-based models for saliency detection have attracted research
...
In this paper, we propose to learn an Unsupervised Single Object Tracker...
Lossy compression algorithms are typically designed to achieve the lowes...
The multiplicative structure of parameters and input data in the first l...
Estimating 3D human pose and shape from a single image is highly
under-c...
In an attempt to better understand structural benefits and generalizatio...
Adversarial attack arises due to the vulnerability of deep neural networ...
The explosive growth of image data facilitates the fast development of i...
A new understanding of adversarial examples and adversarial robustness i...
Automatic self-diagnosis provides low-cost and accessible healthcare via...
The emerging vision-and-language navigation (VLN) problem aims at learni...
Active learning aims to address the paucity of labeled data by finding t...
It is not clear yet why ADAM-alike adaptive gradient algorithms suffer f...
Graph neural networks (GNN) have shown great success in learning from
gr...
Infrared target tracking plays an important role in both civil and milit...
There are two main lines of research on visual question answering (VQA):...
The purpose of this article is to review the achievements made in the la...
A simple approach is proposed to obtain complexity controls for neural
n...
The dynamic behavior of RMSprop and Adam algorithms is studied through a...
The random feature model exhibits a kind of resonance behavior when the
...
Single image deraining regards an input image as a fusion of a backgroun...
The advancement of visual tracking has continuously been brought by deep...
While deep convolutional neural networks (CNNs) are vulnerable to advers...
Visual Question Answering (VQA) has achieved great success thanks to the...
A numerical and phenomenological study of the gradient descent (GD) algo...
The advanced magnetic resonance (MR) image reconstructions such as the
c...
Deep generative models often perform poorly in real-world applications d...