In gastroenteroscopy image analysis, numerous CADs demonstrate that deep learning aids doctors' diagnosis. The shapes and sizes of the lesions are varied. And in the clinic, the dataset appears to be data imbalanc...
In gastroenteroscopy image analysis, numerous CADs demonstrate that deep learning aids doctors' diagnosis. The shapes and sizes of the lesions are varied. And in the clinic, the dataset appears to be data imbalanced. However, existing methods directly classify by texture and ignore lesions with various shapes and sizes. To address the issue above, we propose a deep neural network, which consists of multi-scale feature extraction, contrastive feature learning and a multi-scale feature fusion module. We train the contrastive feature learning module and multi-scale feature fusion module simultaneously to alleviate the issue of data distribution differences. Thus, the proposed network can better identify various categories. Extensive experiments on the Hyper Kvasir dataset show that the proposed Hybrid-M2CL outperforms the benchmark proposed by the dataset with 5.0% Macro Precision, 3.3% Macro Recall, 3.4% Macro F1-score, 3.3% Micro Precision, 3.6% MCC. In addition, it outperforms the SOTA by 1.1% Macro F1-score, 2.6% MCC, and 2.0% B-ACC.
Many existing facial action units (AUs) recognition approaches often enhance the AU representation by combining local features from multiple independent branches, each corresponding to a different AU. However, such mu...
详细信息
ISBN:
(纸本)9781665431774
Many existing facial action units (AUs) recognition approaches often enhance the AU representation by combining local features from multiple independent branches, each corresponding to a different AU. However, such multi-branch combination-based methods usually neglect potential mutual assistance and exclusion relationship between AU branches or simply employ a pre-defined and fixed knowledge-graph as a prior. In addition, extracting features from pre-defined AU regions of regular shapes limits the representation ability. In this paper, we propose a novel Local Global Relational network (LGRNet) for facial AU recognition. LGRNet mainly consists of two novel structures, i.e., a skip-BiLSTM module which models the latent mutual assistance and exclusion relationship among local AU features from multiple branches to enhance the feature robustness, and a feature fusion&refining module which explores the complementarity between local AUs and the whole face in order to refine the local AU features to improve the discriminability. Experiments on the BP4D and DISFA AU datasets show that the proposed approach outperforms the state-of-the-art methods by a large margin.
Graph convolutional networks gain remarkable success in semi-supervised learning on graph-structured data. The key to graph-based semi-supervised learning is capturing the smoothness of labels or features over nodes e...
详细信息
Multi-stage ranking pipelines have been a practical solution in modern search systems, where the first-stage retrieval is to return a subset of candidate documents, and latter stages attempt to re-rank those candidate...
详细信息
In the field of mechanical manufacturing, rolling bearings are important core components. To achieve stable rotational operation, improving diagnostic accuracy has become an urgent issue. In this paper, a hybrid fault...
详细信息
This paper presents a novel controller design for dealing with the challenges posed by Denial-of-Service (DoS) attacks in the context of Artificial Intelligence of Things (AIOT). The proposed design employs a predicti...
This paper presents a novel controller design for dealing with the challenges posed by Denial-of-Service (DoS) attacks in the context of Artificial Intelligence of Things (AIOT). The proposed design employs a prediction approach that utilizes uniform quantization for both encoder and decoder, aiming to effectively mitigate data loss during data transmission in scenarios involving DoS attacks. Furthermore, the study investigates the application of dynamic system design as a means to enhance the security defense mechanisms of AIOT, consequently bolstering the overall resilience of the system. The effectiveness and practicality of the proposed approach are validated through comprehensive simulations, providing empirical evidence of its efficacy.
Modularity serves as an omnipresent paradigm across the spectrum of natural phenomena, societal constructs, and human pursuits, spanning from biological systems to corporate hierarchies and further. Within the realm o...
详细信息
We participated in the Deep Learning Track at TREC 2019. We built a ranking system which combines a search component based on BM25 and a semantic matching component using pretraining knowledge. Our best run achieves N...
详细信息
Social medial become our public ways to share our information in our lives. Crisis management via social medial is becoming indispensable for its tremendous information. While deep learning shows surprising outcome in...
详细信息
This paper describes our work in the background linking task and entity ranking task in TREC 2018 News Track. We explore four methods in background linking task and two methods in entity ranking task. All of our metho...
暂无评论