Zobrazeno 1 - 10
of 31
pro vyhledávání: '"Zhong, Qiaoyong"'
The knowledge replay technique has been widely used in many tasks such as continual learning and continuous domain adaptation. The key lies in how to effectively encode the knowledge extracted from previous data and replay them during current trainin
Externí odkaz:
http://arxiv.org/abs/2205.11126
Exemplar-free incremental learning is extremely challenging due to inaccessibility of data from old tasks. In this paper, we attempt to exploit the knowledge encoded in a previously trained classification model to handle the catastrophic forgetting p
Externí odkaz:
http://arxiv.org/abs/2205.11071
In the context of skeleton-based action recognition, graph convolutional networks (GCNs) have been rapidly developed, whereas convolutional neural networks (CNNs) have received less attention. One reason is that CNNs are considered poor in modeling t
Externí odkaz:
http://arxiv.org/abs/2112.04178
Reconstruction-based methods play an important role in unsupervised anomaly detection in images. Ideally, we expect a perfect reconstruction for normal samples and poor reconstruction for abnormal samples. Since the generalizability of deep neural ne
Externí odkaz:
http://arxiv.org/abs/2107.13118
Object detection involves two sub-tasks, i.e. localizing objects in an image and classifying them into various categories. For existing CNN-based detectors, we notice the widespread divergence between localization and classification, which leads to d
Externí odkaz:
http://arxiv.org/abs/2103.08958
Graph Convolutional Networks (GCNs) have attracted increasing interests for the task of skeleton-based action recognition. The key lies in the design of the graph structure, which encodes skeleton topology information. In this paper, we propose Dynam
Externí odkaz:
http://arxiv.org/abs/2007.14690
Autor:
She, Qi, Feng, Fan, Liu, Qi, Chan, Rosa H. M., Hao, Xinyue, Lan, Chuanlin, Yang, Qihan, Lomonaco, Vincenzo, Parisi, German I., Bae, Heechul, Brophy, Eoin, Chen, Baoquan, Graffieti, Gabriele, Goel, Vidit, Han, Hyonyoung, Kanagarajah, Sathursan, Kumar, Somesh, Lam, Siew-Kei, Lam, Tin Lun, Ma, Liang, Maltoni, Davide, Pellegrini, Lorenzo, Piyasena, Duvindu, Pu, Shiliang, Sheet, Debdoot, Song, Soonyong, Son, Youngsung, Wang, Zhengwei, Ward, Tomas E., Wu, Jianwen, Wu, Meiqing, Xie, Di, Xu, Yangsheng, Yang, Lin, Zhong, Qiaoyong, Zhou, Liguang
This report summarizes IROS 2019-Lifelong Robotic Vision Competition (Lifelong Object Recognition Challenge) with methods and results from the top $8$ finalists (out of over~$150$ teams). The competition dataset (L)ifel(O)ng (R)obotic V(IS)ion (OpenL
Externí odkaz:
http://arxiv.org/abs/2004.14774
Spatio-temporal feature learning is of central importance for action recognition in videos. Existing deep neural network models either learn spatial and temporal features independently (C2D) or jointly with unconstrained parameters (C3D). In this pap
Externí odkaz:
http://arxiv.org/abs/1903.01197
Person re-identification (ReID) aims to match people across multiple non-overlapping video cameras deployed at different locations. To address this challenging problem, many metric learning approaches have been proposed, among which triplet loss is o
Externí odkaz:
http://arxiv.org/abs/1812.06576
Skeleton-based human action recognition has recently drawn increasing attentions with the availability of large-scale skeleton datasets. The most crucial factors for this task lie in two aspects: the intra-frame representation for joint co-occurrence
Externí odkaz:
http://arxiv.org/abs/1804.06055