In this paper, we show that recent advances in video representation lear...
Unsupervised object-centric learning methods allow the partitioning of s...
Amodal object segmentation is a challenging task that involves segmentin...
Recent research understands the residual networks from a new perspective...
In this work, we propose SAM3D, a novel framework that is able to predic...
Residual networks have shown great success and become indispensable in r...
Although the variational autoencoder (VAE) and its conditional extension...
Layout-to-image generation refers to the task of synthesizing photo-real...
Neural Architecture Search has attracted increasing attention in recent
...
We propose a novel approach to self-supervised learning of point cloud
r...
Masked Modeling (MM) has demonstrated widespread success in various visi...
The pretraining-finetuning paradigm has demonstrated great success in NL...
Despite the tremendous progress of Masked Autoencoders (MAE) in developi...
This paper presents an approach that reconstructs a hand-held object fro...
Previous top-performing methods for 3D instance segmentation often maint...
Data augmentations are important in training high-performance 3D object
...
Amodal perception requires inferring the full shape of an object that is...
Long-tail distribution is widely spread in real-world applications. Due ...
Humans naturally decompose their environment into entities at the approp...
The current state-of-the-art methods in 3D instance segmentation typical...
We present ARCH++, an image-based method to reconstruct 3D avatars with
...
Recognizing and localizing objects in the 3D space is a crucial ability ...
We propose an approach to instance segmentation from 3D point clouds bas...
We propose a hierarchical graph neural network (GNN) model that learns h...
Scene flow in 3D point clouds plays an important role in understanding
d...
End-to-end text-spotting, which aims to integrate detection and recognit...
Previous top-performing approaches for point cloud instance segmentation...
We propose Geo-PIFu, a method to recover a 3D mesh from a monocular colo...
In computer vision, object detection is one of most important tasks, whi...
Deep learning usually achieves the best results with complete supervisio...
While image classification models have recently continued to advance, mo...
Scene text detection and recognition has received increasing research
at...
3D point cloud semantic and instance segmentation is crucial and fundame...
A collection of approaches based on graph convolutional networks have pr...
Omnidirectional scene text detection has received increasing research
at...
We describe a policy learning approach to map visual inputs to driving
c...
We present GluonCV and GluonNLP, the deep learning toolkits for computer...
With an increasing demand for training powers for deep learning algorith...
We propose a fully convolutional one-stage object detector (FCOS) to sol...
Both accuracy and efficiency are of significant importance to the task o...
Recent semantic segmentation methods exploit encoder-decoder architectur...
Comparing with enormous research achievements targeting better image
cla...
We present a method to infer 3D pose and shape of vehicles from a single...
Graphics Interchange Format (GIF) is a highly portable graphics format t...
Surface-based geodesic topology provides strong cues for object semantic...
Much of the recent progress made in image classification research can be...
With the rapid growth of fashion-focused social networks and online shop...
Text detection and recognition in natural images have long been consider...
We propose a novel Connectionist Text Proposal Network (CTPN) that accur...
We introduce a new top-down pipeline for scene text detection. We propos...