Search

The physical attack has been regarded as a kind of threat against real-world computer vision systems. Still, many existing defense methods are only useful for small perturbations attacks and can't detect physical attacks effectively. In this paper, we propose a random-patch based defense strategy to robustly detect physical attacks for Face Recognition System (FRS). Different from mainstream defense methods which focus on building complex deep neural networks (DNN) to achieve high recognition rate on attacks, we introduce a patch based defense strategy to a standard DNN aiming to obtain robust detection models. Extensive experimental results on the employed datasets show the superiority of the proposed defense method on detecting white-box attacks and adaptive attacks which attack both FRS and the defense method. Additionally, due to the simpleness yet robustness of our method, it can be easily applied to the real world face recognition system and extended to other defense methods to boost the detection performance.

Data in real-world object detection often exhibits the long-tailed distribution. Existing solutions tackle this problem by mitigating the competition between the head and tail categories. However, due to the scarcity of training samples, tail categories are still unable to learn discriminative representations. Bringing more data into the training may alleviate the problem, but collecting instance-level annotations is an excruciating task. In contrast, image-level annotations are easily accessible but not fully exploited. In this paper, we propose a novel framework CLIS (multi-task Collaborative Learning with Image-level Supervision), which leverage image-level supervision to enhance the detection ability in a multi-task collaborative way. Specifically, there are an object detection task (consisting of an instance-classification task and a localization task) and an image-classification task in our framework, responsible for utilizing the two types of supervision. Different tasks are trained collaboratively by three key designs: (1) task-specialized sub-networks that learn specific representations of different tasks without feature entanglement. (2) a siamese sub-network for the image-classification task that shares its knowledge with the instance-classification task, resulting in feature enrichment of detectors. (3) a contrastive learning regularization that maintains representation consistency, bridging feature gaps of different supervision. Extensive experiments are conducted on the challenging LVIS dataset. Without sophisticated loss engineering, CLIS achieves an overall AP of 31.1 with 10.1 point improvement on tail categories, establishing a new state-of-the-art. Code will be at https://github.com/waveboo/CLIS.

Despite the recent success of long-tailed object detection, almost all long-tailed object detectors are developed based on the two-stage paradigm. In practice, one-stage detectors are more prevalent in the industry because they have a simple and fast pipeline that is easy to deploy. However, in the long-tailed scenario, this line of work has not been explored so far. In this paper, we investigate whether one-stage detectors can perform well in this case. We discover the primary obstacle that prevents one-stage detectors from achieving excellent performance is: categories suffer from different degrees of positive-negative imbalance problems under the long-tailed data distribution. The conventional focal loss balances the training process with the same modulating factor for all categories, thus failing to handle the long-tailed problem. To address this issue, we propose the Equalized Focal Loss (EFL) that rebalances the loss contribution of positive and negative samples of different categories independently according to their imbalance degrees. Specifically, EFL adopts a category-relevant modulating factor which can be adjusted dynamically by the training status of different categories. Extensive experiments conducted on the challenging LVIS v1 benchmark demonstrate the effectiveness of our proposed method. With an end-to-end training pipeline, EFL achieves 29.2% in terms of overall AP and obtains significant performance improvements on rare categories, surpassing all existing state-of-the-art methods. The code is available at https://github.com/ModelTC/EOD., Comment: Accepted by the IEEE/CVF Computer Vision and Pattern Recognition Conference (CVPR) 2022

A modified rock mass classification system for TBM tunnels is proposed based on the hydropower rock mass classification (HC) method used in China. Rock mass properties and TBM performance data from three tunnelling projects were used for regression analysis. The proposed method considers the rock mass boreability and stability, and estimates the penetration per revolution, utilisation, and advance rate of a TBM under various rock mass conditions. The influence of adverse geological conditions (rockbursts, squeezing, faults, highly abrasive rock, and water inrush) on TBM performance is also considered. The classification system can be used to evaluate the feasibility of TBM construction and serve as the basis for selection of TBM types and design parameters in the pre-construction phase. It can also be applied to estimate TBM performance and guide tunnel support for a given ground condition in the design phase, and to optimise TBM operational parameters during the construction phase. A 7.9 m diameter section of the YE water conveyance tunnel, which was driven through granite in Xinjiang, is used to illustrate the application of the modified rock mass classification system., A modified rock mass classification system for TBM tunnels is proposed based on the hydropower rock mass classification (HC) method used in China. Rock mass properties and TBM performance data from three tunnelling projects were used for regression analysis. The proposed method considers the rock mass boreability and stability, and estimates the penetration per revolution, utilisation, and advance rate of a TBM under various rock mass conditions. The influence of adverse geological conditions (rockbursts, squeezing, faults, highly abrasive rock, and water inrush) on TBM performance is also considered. The classification system can be used to evaluate the feasibility of TBM construction and serve as the basis for selection of TBM types and design parameters in the pre-construction phase. It can also be applied to estimate TBM performance and guide tunnel support for a given ground condition in the design phase, and to optimise TBM operational parameters during the construction phase. A 7.9 m diameter section of the YE water conveyance tunnel, which was driven through granite in Xinjiang, is used to illustrate the application of the modified rock mass classification system.

Hui B, Xu Y, Zhao B, et al. Onco Targets Ther. 2019;12:1043–1057. The Editor and Publisher of OncoTargets and Therapy wish to retract the published article. Concerns were raised over alleged image duplication between Figures 2F and 6E. Specifically: Figure 2F, panel AsPC-1 Empty vector appears to have been duplicated with Figure 6E, panel BxPC-3 si-NC. The author responded to our queries but were unable to provide a satisfactory explanation for the alleged duplication and could not provide the original data for the study. The Editor requested for the article to be retracted and the authors agreed with this decision. The authors wish to apologise for this error. Our decision-making was informed by our policy on publishing ethics and integrity and the COPE guidelines on retraction. The retracted article will remain online to maintain the scholarly record, but it will be digitally watermarked on each page as “Retracted”. This retraction relates to this paper

The application of deep learning to medical image segmentation has been hampered due to the lack of abundant pixel-level annotated data. Few-shot Semantic Segmentation (FSS) is a promising strategy for breaking the deadlock. However, a high-performing FSS model still requires sufficient pixel-level annotated classes for training to avoid overfitting, which leads to its performance bottleneck in medical image segmentation due to the unmet need for annotations. Thus, semi-supervised FSS for medical images is accordingly proposed to utilize unlabeled data for further performance improvement. Nevertheless, existing semi-supervised FSS methods has two obvious defects: (1) neglecting the relationship between the labeled and unlabeled data; (2) using unlabeled data directly for end-to-end training leads to degenerated representation learning. To address these problems, we propose a novel semi-supervised FSS framework for medical image segmentation. The proposed framework employs Poisson learning for modeling data relationship and propagating supervision signals, and Spatial Consistency Calibration for encouraging the model to learn more coherent representations. In this process, unlabeled samples do not involve in end-to-end training, but provide supervisory information for query image segmentation through graph-based learning. We conduct extensive experiments on three medical image segmentation datasets (i.e. ISIC skin lesion segmentation, abdominal organs segmentation for MRI and abdominal organs segmentation for CT) to demonstrate the state-of-the-art performance and broad applicability of the proposed framework., Comment: Accepted by 2021 IEEE International Conference on Bioinformatics and Biomedicine (BIBM 2021)

Coarse-to-fine models and cascade segmentation architectures are widely adopted to solve the problem of large scale variations in medical image segmentation. However, those methods have two primary limitations: the first-stage segmentation becomes a performance bottleneck; the lack of overall differentiability makes the training process of two stages asynchronous and inconsistent. In this paper, we propose a differentiable two-stage network architecture to tackle these problems. In the first stage, a localization network (L-Net) locates Regions of Interest (RoIs) in a detection fashion; in the second stage, a segmentation network (S-Net) performs fine segmentation on the recalibrated RoIs; a RoI recalibration module between L-Net and S-Net eliminating the inconsistencies. Experimental results on the public dataset show that our method outperforms state-of-the-art coarse-to-fine models with negligible computation overheads.

Automatic segmentation of the prostate cancer from the multi-modal magnetic resonance images is of critical importance for the initial staging and prognosis of patients. However, how to use the multi-modal image features more efficiently is still a challenging problem in the field of medical image segmentation. In this paper, we develop a cross-modal self-attention distillation network by fully exploiting the encoded information of the intermediate layers from different modalities, and the extracted attention maps of different modalities enable the model to transfer the significant spatial information with more details. Moreover, a novel spatial correlated feature fusion module is further employed for learning more complementary correlation and non-linear information of different modality images. We evaluate our model in five-fold cross-validation on 358 MRI with biopsy confirmed. Extensive experiment results demonstrate that our proposed network achieves state-of-the-art performance., Comment: 2020 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

Kang He,1,* Yajing Wang,1,* Yuejiao Zhong,2 Xiaohua Pan,1 Lixiang Si,1 Jianwei Lu1 1The Department of Oncology, The Affiliated Cancer Hospital of Nanjing Medical University, and Jiangsu Cancer Hospital and Jiangsu Institute of Cancer Research, Nanjing, People’s Republic of China; 2The Department of Oncology, Jiangsu Cancer Hospital, and the Affiliated Cancer Hospital of Nanjing Medical University and Jiangsu Institute of Cancer Research, Nanjing, People’s Republic of China*These authors contributed equally to this workCorrespondence: Jianwei LuThe Affiliated Cancer Hospital of Nanjing Medical University and Jiangsu Cancer Hospital and Jiangsu Institute of Cancer Research, Nanjing, People’s Republic of ChinaEmail lujw@medmail.com.cnObjective: To investigate the connection between mutant KRAS/NRAS/BRAF and clinicopathological characteristics in therapy-naïve synchronous metastatic colorectal cancer (mCRC) in Chinese populations when compared with all wild type (KRAS/NRAS/BRAF wild type).Patients and Methods: A total of 200 patients with therapy-naïve synchronous mCRC (TNM stage: TanyNanyM1) were retrospectively collected as study objects. Primary tumor tissues from 200 mCRC patients were analyzed through next-generation sequencing panel to assess the mutated regions of KRAS/NRAS/BRAF.Results: The distribution frequency of gene mutation in our study was 41% KRAS, 4% NRAS, 11.5% BRAF, 0.5% both KRAS and BRAF. Tumors with any gene mutations (any gene mutations in KRAS/NRAS/BRAF), KRAS and KRAS codon 12 mutation were more likely to be located in right-sided colon (P=0.007, P=0.008, P=0.026, respectively). For metastasis, tumors with any gene mutations, KRAS and KRAS codon 12 mutation were significantly correlated with peritoneal metastasis (P=0.019, P=0.017, P=0.014, respectively), liver-peritoneum metastases (P=0.004, P=0.003, P=0.002, respectively) and multi-organ metastases (P=0.002, P=0.008, P=0.001, respectively). Tumor

Xuan Pan,* Yajing Wang,* Chenchen Li,* Zhaofei Zhou, Yuejiao Zhong, Jifeng Feng, Jianwei Lu Department of Medical Oncology, The Affiliated Cancer Hospital of Nanjing Medical University, Jiangsu Cancer Hospital, Jiangsu Institute of Cancer Research, Nanjing, Jiangsu 210009, People’s Republic of China*These authors contributed equally to this workCorrespondence: Jifeng Feng; Jianwei LuDepartment of Medical Oncology, The Affiliated Cancer Hospital of Nanjing Medical University, Jiangsu Cancer Hospital, Jiangsu Institute of Cancer Research, 42 Baiziting Road, Nanjing, Jiangsu 210009, People’s Republic of ChinaEmail fjfvip@126.com; lujw@medmail.com.cnIntroduction: Gastric cancer is highly heterogeneous both clinically and pathologically and is one of the leading causes of cancer-related deaths worldwide. Genomic coverage variations, also known as copy number variations (CNVs), play a critical role in the carcinogenesis of gastric cancer. Many studies have demonstrated that DNA CNVs are important factors affecting the expression of protein-encoding genes in the gastric cancer genome.Methods: Thirty gastric cancer patients from a Chinese population were enrolled. Genomic DNA was extracted from gastric cancer tissue and matched adjacent non-cancerous tissue from each patient. A panel of 1,021 genes including 3300 exons was designed and subjected to next-generation sequencing. Copy numbers of each gene and exon were calculated for each tissue. Coverage variations between gastric cancer tissue and matched adjacent non-cancerous tissue were also calculated, and we examined the correlation between overall survival of patients and coverage variation type for each exon.Results: DNA from cancerous tissue and corresponding adjacent non-cancerous tissue were significantly different with respect to the pattern of gene copy number. Exon copy numbers were highly consistent among non-cancerous samples and confirmed that non-cancerous tissue contain diploid genomes. I

This paper has been retracted.Bingqing Hui, 1,2,* Yetao Xu, 3,* Benpeng Zhao, 4,* Hao Ji, 1,2 Zhonghua Ma, 1,2 Shufen Xu, 1,2 ZhenYu He, 1,5 Keming Wang, 1,2 Jianwei Lu 6 1Department of Oncology, Second Affiliated Hospital, Nanjing Medical University, Nanjing 210000, Jiangsu, China; 2Department of Oncology, Second Clinical Medical College of Nanjing Medical University, Nanjing 210000, Jiangsu, China; 3Department of Obstetrics and Gynecology, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, Jiangsu, China; 4Basic Medicine Faculty of Shanghai Jiaotong University, Core Facility of Basic Medical Sciences, Shanghai 200000, China; 5Department of General Surgery, Second Affiliated Hospital, Nanjing Medical University, Nanjing 210000, Jiangsu, China; 6Department of Medical Oncology, Jiangsu Cancer Hospital, Jiangsu Institute of Cancer Research, Nanjing Medical University Affiliated Cancer Hospital, Nanjing 210000, Jiangsu, China *These authors contributed equally to this work Background: Long noncoding RNAs (lncRNAs) are involved in various human diseases, including cancers. However, their mechanisms remain undocumented. We investigated alterations in lncRNA that may be related to pancreatic cancer (PC) through analysis of microarray data. Methods: In the present study, quantitative real-time PCR analysis was used to examine the expression of taurine upregulated 1 (TUG1) in PC tissue samples and PC cell lines. In PC cell lines, MTT assays, colony formation assays, and flow cytometry were used to investigate the effects of TUG1 on proliferation, cell cycle regulation, and apoptosis. Moreover, we established a xenograft model to assess the effect of TUG1 on tumor growth in vivo. The molecular mechanism of potential target genes was detected through nuclear separation experiments, RNA immunoprecipitation (RIP), chromatin immunoprecipitation assays (ChIP), and other experimental methods. Results: The findings suggest that the abnormally high expressi

Spammer detection on social network is a challenging problem. The rigid anti-spam rules have resulted in emergence of "smart" spammers. They resemble legitimate users who are difficult to identify. In this paper, we present a novel spammer classification approach based on Latent Dirichlet Allocation(LDA), a topic model. Our approach extracts both the local and the global information of topic distribution patterns, which capture the essence of spamming. Tested on one benchmark dataset and one self-collected dataset, our proposed method outperforms other state-of-the-art methods in terms of averaged F1-score., Comment: NAACL-HLT 2016, Student Research Workshop

Developing 3D palmprint recognition systems has recently begun to draw attention of researchers. Compared with its 2D counterpart, 3D palmprint has several unique merits. However, most of the existing 3D palmprint matching methods are designed for one-to-one verification and they are not efficient to cope with the one-to-many identification case. In this paper, we fill this gap by proposing a collaborative representation (CR) based framework with l1-norm or l2-norm regularizations for 3D palmprint identification. The effects of different regularization terms have been evaluated in experiments. To use the CR-based classification framework, one key issue is how to extract feature vectors. To this end, we propose a block-wise statistics based feature extraction scheme. We divide a 3D palmprint ROI into uniform blocks and extract a histogram of surface types from each block; histograms from all blocks are then concatenated to form a feature vector. Such feature vectors are highly discriminative and are robust to mere misalignment. Experiments demonstrate that the proposed CR-based framework with an l2-norm regularization term can achieve much better recognition accuracy than the other methods. More importantly, its computational complexity is extremely low, making it quite suitable for the large-scale identification application. Source codes are available at http://sse.tongji.edu.cn/linzhang/cr3dpalm/cr3dpalm.htm. © 2014 IEEE.

Developing 3D palmprint recognition systems has recently begun to draw attention of researchers. Compared with its 2D counterpart, 3D palmprint has several unique merits. However, most of the existing 3D palmprint matching methods are designed for one-to-one verification and they are not efficient to cope with the one-to-many identification case. In this paper, we fill this gap by proposing a collaborative representation (CR) based framework with l1-norm or l2-norm regularizations for 3D palmprint identification. The effects of different regularization terms have been evaluated in experiments. To use the CR-based classification framework, one key issue is how to extract feature vectors. To this end, we propose a block-wise statistics based feature extraction scheme. We divide a 3D palmprint ROI into uniform blocks and extract a histogram of surface types from each block; histograms from all blocks are then concatenated to form a feature vector. Such feature vectors are highly discriminative and are robust to mere misalignment. Experiments demonstrate that the proposed CR-based framework with an l2-norm regularization term can achieve much better recognition accuracy than the other methods. More importantly, its computational complexity is extremely low, making it quite suitable for the large-scale identification application. Source codes are available at http://sse.tongji.edu.cn/linzhang/cr3dpalm/cr3dpalm.htm. © 2014 IEEE.

Developing 3D palmprint recognition systems has recently begun to draw attention of researchers. Compared with its 2D counterpart, 3D palmprint has several unique merits. However, most of the existing 3D palmprint matching methods are designed for one-to-one verification and they are not efficient to cope with the one-to-many identification case. In this paper, we fill this gap by proposing a collaborative representation (CR) based framework with l1-norm or l2-norm regularizations for 3D palmprint identification. The effects of different regularization terms have been evaluated in experiments. To use the CR-based classification framework, one key issue is how to extract feature vectors. To this end, we propose a block-wise statistics based feature extraction scheme. We divide a 3D palmprint ROI into uniform blocks and extract a histogram of surface types from each block; histograms from all blocks are then concatenated to form a feature vector. Such feature vectors are highly discriminative and are robust to mere misalignment. Experiments demonstrate that the proposed CR-based framework with an l2-norm regularization term can achieve much better recognition accuracy than the other methods. More importantly, its computational complexity is extremely low, making it quite suitable for the large-scale identification application. Source codes are available at http://sse.tongji.edu.cn/linzhang/cr3dpalm/cr3dpalm.htm. © 2014 IEEE.