ORCID Profile
0000-0001-7114-5854
Current Organisations
Zhejiang University
,
Greenland Institute of Natural Resources
Does something not look right? The information on this page has been harvested from data sources that may not be up to date. We continue to work with information providers to improve coverage and quality. To report an issue, use the Feedback Form.
In Research Link Australia (RLA), "Research Topics" refer to ANZSRC FOR and SEO codes. These topics are either sourced from ANZSRC FOR and SEO codes listed in researchers' related grants or generated by a large language model (LLM) based on their publications.
Pattern Recognition and Data Mining | Artificial Intelligence and Image Processing | Database Management | Computer Vision | Artificial Intelligence and Image Processing not elsewhere classified | Information Systems | Neural, Evolutionary and Fuzzy Computation | Multimedia Programming
Information Processing Services (incl. Data Entry and Capture) | Electronic Information Storage and Retrieval Services | Film and Video Services (excl. Animation and Computer Generated Imagery) | Media Services not elsewhere classified | Application Tools and System Utilities | Expanding Knowledge in the Information and Computing Sciences | Health Policy Evaluation |
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 03-2021
Publisher: Association for Computing Machinery (ACM)
Date: 10-10-2018
DOI: 10.1145/3243316
Abstract: The superiority of deeply learned pedestrian representations has been reported in very recent literature of person re-identification (re-ID). In this article, we consider the more pragmatic issue of learning a deep feature with no or only a few labels. We propose a progressive unsupervised learning (PUL) method to transfer pretrained deep representations to unseen domains. Our method is easy to implement and can be viewed as an effective baseline for unsupervised re-ID feature learning. Specifically, PUL iterates between (1) pedestrian clustering and (2) fine-tuning of the convolutional neural network (CNN) to improve the initialization model trained on the irrelevant labeled dataset. Since the clustering results can be very noisy, we add a selection operation between the clustering and fine-tuning. At the beginning, when the model is weak, CNN is fine-tuned on a small amount of reliable ex les that locate near to cluster centroids in the feature space. As the model becomes stronger, in subsequent iterations, more images are being adaptively selected as CNN training s les. Progressively, pedestrian clustering and the CNN model are improved simultaneously until algorithm convergence. This process is naturally formulated as self-paced learning. We then point out promising directions that may lead to further improvement. Extensive experiments on three large-scale re-ID datasets demonstrate that PUL outputs discriminative features that improve the re-ID accuracy. Our code has been released at ehefan/Unsupervised-Person-Re-identification-Clustering-and-Fine-tuning.
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 2019
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 02-2014
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 2021
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 03-2019
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 10-2019
Publisher: Frontiers Media SA
Date: 25-06-2021
Abstract: Southern Ocean ecosystems are globally important and vulnerable to global drivers of change, yet they remain challenging to study. Fish and squid make up a significant portion of the biomass within the Southern Ocean, filling key roles in food webs from forage to mid-trophic species and top predators. They comprise a erse array of species uniquely adapted to the extreme habitats of the region. Adaptations such as antifreeze glycoproteins, lipid-retention, extended larval phases, delayed senescence, and energy-conserving life strategies equip Antarctic fish and squid to withstand the dark winters and yearlong subzero temperatures experienced in much of the Southern Ocean. In addition to krill exploitation, the comparatively high commercial value of Antarctic fish, particularly the lucrative toothfish, drives fisheries interests, which has included illegal fishing. Uncertainty about the population dynamics of target species and ecosystem structure and function more broadly has necessitated a precautionary, ecosystem approach to managing these stocks and enabling the recovery of depleted species. Fisheries currently remain the major local driver of change in Southern Ocean fish productivity, but global climate change presents an even greater challenge to assessing future changes. Parts of the Southern Ocean are experiencing ocean-warming, such as the West Antarctic Peninsula, while other areas, such as the Ross Sea shelf, have undergone cooling in recent years. These trends are expected to result in a redistribution of species based on their tolerances to different temperature regimes. Climate variability may impair the migratory response of these species to environmental change, while imposing increased pressures on recruitment. Fisheries and climate change, coupled with related local and global drivers such as pollution and sea ice change, have the potential to produce synergistic impacts that compound the risks to Antarctic fish and squid species. The uncertainty surrounding how different species will respond to these challenges, given their varying life histories, environmental dependencies, and resiliencies, necessitates regular assessment to inform conservation and management decisions. Urgent attention is needed to determine whether the current management strategies are suitably precautionary to achieve conservation objectives in light of the impending changes to the ecosystem.
Publisher: Springer International Publishing
Date: 2018
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 05-2018
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 04-2018
Publisher: Association for Computing Machinery (ACM)
Date: 22-05-2020
DOI: 10.1145/3383184
Abstract: Matching images and sentences demands a fine understanding of both modalities. In this article, we propose a new system to discriminatively embed the image and text to a shared visual-textual space. In this field, most existing works apply the ranking loss to pull the positive image/text pairs close and push the negative pairs apart from each other. However, directly deploying the ranking loss on heterogeneous features (i.e., text and image features) is less effective, because it is hard to find appropriate triplets at the beginning. So the naive way of using the ranking loss may compromise the network from learning inter-modal relationship. To address this problem, we propose the instance loss, which explicitly considers the intra-modal data distribution. It is based on an unsupervised assumption that each image/text group can be viewed as a class. So the network can learn the fine granularity from every image/text group. The experiment shows that the instance loss offers better weight initialization for the ranking loss, so that more discriminative embeddings can be learned. Besides, existing works usually apply the off-the-shelf features, i.e., word2vec and fixed visual feature. So in a minor contribution, this article constructs an end-to-end dual-path convolutional network to learn the image and text representations. End-to-end learning allows the system to directly learn from the data and fully utilize the supervision. On two generic retrieval datasets (Flickr30k and MSCOCO), experiments demonstrate that our method yields competitive accuracy compared to state-of-the-art methods. Moreover, in language-based person retrieval, we improve the state of the art by a large margin. The code has been made publicly available.
Publisher: IEEE
Date: 07-2017
Publisher: Elsevier BV
Date: 2016
Publisher: IEEE
Date: 12-2013
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 07-2019
Publisher: Springer International Publishing
Date: 2018
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 2023
Publisher: IEEE
Date: 10-2017
Publisher: Association for Computing Machinery (ACM)
Date: 13-12-2017
DOI: 10.1145/3159171
Abstract: In this article, we revisit two popular convolutional neural networks in person re-identification (re-ID): verification and identification models. The two models have their respective advantages and limitations due to different loss functions. Here, we shed light on how to combine the two models to learn more discriminative pedestrian descriptors. Specifically, we propose a Siamese network that simultaneously computes the identification loss and verification loss. Given a pair of training images, the network predicts the identities of the two input images and whether they belong to the same identity. Our network learns a discriminative embedding and a similarity measurement at the same time, thus taking full usage of the re-ID annotations. Our method can be easily applied on different pretrained networks. Albeit simple, the learned embedding improves the state-of-the-art performance on two public person re-ID benchmarks. Further, we show that our architecture can also be applied to image retrieval. The code is available at ayumi/2016_person_re-ID.
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 07-2016
Publisher: IEEE
Date: 06-2012
Publisher: Springer Science and Business Media LLC
Date: 04-2019
DOI: 10.1038/S41586-019-1086-6
Abstract: Lightning is a dangerous yet poorly understood natural phenomenon. Lightning forms a network of plasma channels propagating away from the initiation point with both positively and negatively charged ends-called positive and negative leaders
Publisher: Springer Science and Business Media LLC
Date: 28-03-2023
Publisher: Springer Science and Business Media LLC
Date: 20-12-2022
Publisher: Springer Nature Switzerland
Date: 2022
Publisher: IEEE
Date: 06-2014
DOI: 10.1109/CVPR.2014.20
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 09-2019
Publisher: Association for Computing Machinery (ACM)
Date: 07-02-2019
DOI: 10.1145/3300939
Abstract: Performing direct matching among different modalities (like image and text) can benefit many tasks in computer vision, multimedia, information retrieval, and information fusion. Most of existing works focus on class-level image-text matching, called cross-modal retrieval , which attempts to propose a uniform model for matching images with all types of texts, for ex le, tags, sentences, and articles (long texts). Although cross-model retrieval alleviates the heterogeneous gap among visual and textual information, it can provide only a rough correspondence between two modalities. In this article, we propose a more precise image-text embedding method, image-sentence matching, which can provide heterogeneous matching in the instance level. The key issue for image-text embedding is how to make the distributions of the two modalities consistent in the embedding space. To address this problem, some previous works on the cross-model retrieval task have attempted to pull close their distributions by employing adversarial learning. However, the effectiveness of adversarial learning on image-sentence matching has not been proved and there is still not an effective method. Inspired by previous works, we propose to learn a modality-invariant image-text embedding for image-sentence matching by involving adversarial learning. On top of the triplet loss--based baseline, we design a modality classification network with an adversarial loss, which classifies an embedding into either the image or text modality. In addition, the multi-stage training procedure is carefully designed so that the proposed network not only imposes the image-text similarity constraints by ground-truth labels, but also enforces the image and text embedding distributions to be similar by adversarial learning. Experiments on two public datasets (Flickr30k and MSCOCO) demonstrate that our method yields stable accuracy improvement over the baseline model and that our results compare favorably to the state-of-the-art methods.
Publisher: Springer International Publishing
Date: 2018
Publisher: ACM
Date: 22-10-2013
Publisher: Springer International Publishing
Date: 2018
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 2022
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 11-2022
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 2023
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 02-2020
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 2021
Start Date: 05-2013
End Date: 12-2016
Amount: $375,000.00
Funder: Australian Research Council
View Funded ActivityStart Date: 07-2016
End Date: 07-2020
Amount: $520,000.00
Funder: Australian Research Council
View Funded ActivityStart Date: 2015
End Date: 12-2018
Amount: $494,300.00
Funder: Australian Research Council
View Funded ActivityStart Date: 05-2020
End Date: 03-2025
Amount: $486,000.00
Funder: Australian Research Council
View Funded ActivityStart Date: 2018
End Date: 12-2021
Amount: $392,884.00
Funder: Australian Research Council
View Funded Activity