A GAN-Based Self-Training Framework for Unsupervised Domain Adaptive Person Re-Identification.

As a crucial task in surveillance and security, person re-identification (re-ID) aims to identify the targeted pedestrians across multiple images captured by non-overlapping cameras. However, existing person re-ID solutions have two main challenges: the lack of pedestrian identification labels in the captured images, and domain shift issue between different domains. A generative adversarial networks (GAN)-based self-training framework with progressive augmentation (SPA) is proposed to obtain the robust features of the unlabeled data from the target domain, according to the preknowledge of the labeled data from the source domain. Specifically, the proposed framework consists of two stages: the style transfer stage (STrans), and self-training stage (STrain). First, the targeted data is complemented by a camera style transfer algorithm in the STrans stage, in which CycleGAN and Siamese Network are integrated to preserve the unsupervised self-similarity (the similarity of the same image between before and after transformation) and domain dissimilarity (the dissimilarity between a transferred source image and the targeted image). Second, clustering and classification are alternately applied to enhance the model performance progressively in the STrain stage, in which both global and local features of the target-domain images are obtained. Compared with the state-of-the-art methods, the proposed method achieves the competitive accuracy on two existing datasets.

EXAM: A Framework of Learning Extreme and Moderate Embeddings for Person Re-ID.

Person re-identification (Re-ID) is challenging due to host of factors: the variety of human positions, difficulties in aligning bounding boxes, and complex backgrounds, among other factors. This paper proposes a new framework called EXAM (EXtreme And Moderate feature embeddings) for Re-ID tasks. This is done using discriminative feature learning, requiring attention-based guidance during training. Here "Extreme" refers to salient human features and "Moderate" refers to common human features. In this framework, these types of embeddings are calculated by global max-pooling and average-pooling operations respectively; and then, jointly supervised by multiple triplet and cross-entropy loss functions. The processes of deducing attention from learned embeddings and discriminative feature learning are incorporated, and benefit from each other in this end-to-end framework. From the comparative experiments and ablation studies, it is shown that the proposed EXAM is effective, and its learned feature representation reaches state-of-the-art performance.

Expression of WIF-1 in inflammatory bowel disease.

The WNT/ß-catenin cellular network has been extensively studied in numerous diseases including inflammatory bowel disease (IBD). IBD is a condition that increases the risk of developing colorectal cancer. WIF-1 is an inhibitory protein that acts by blocking the interactions of WNT with its receptor complex, thus leading to downregulation of end products of this pathway. While WIF-1 has been characterized in several cancers, its relationship with IBD has yet to be elucidated. In this study, the expression of WIF-1 in patients with IBD was analyzed in order to provide insights into the pathophysiology and rationale for alternative therapies. Biopsies of both normal and inflamed colonic mucosa from patients with Crohn's disease or ulcerative colitis were histologically examined for the degree of morphologic changes, immune cell infiltration and presence of WIF-1 through immunohistochemistry. No differences were observed in WIF-1 expression linked to a particular condition, but WIF-1 stain was significantly enhanced in the crypts and lamina propria as inflammation increased in biopsies from patients with both, ulcerative colitis and Crohn's disease. These findings could give guidance to new therapeutic applications of the WNT/ß-catenin system and WIF-1 in IBD.

Serum bilirubin as a prognostic marker in patients with acute decompensated heart failure.

BACKGROUND/AIMS: Several prognostic markers for heart failure (HF) have been determined but the importance of liver function tests (LFTs) remains unknown. The aim of this study was to determine the prognostic significance, if any, of abnormal LFTs in acute decompensated HF. METHODS: All adult patients (> 18 years of age) who were admitted to a community hospital with a diagnosis of acute decompensated HF during the period January 2008 to December 2009 were identified. Exclusion criteria included acute coronary syndrome, active hepatobiliary disease, renal failure (serum creatinine ≥ 2 mg/dL), and malignancy. The primary end point was readmission secondary to acute exacerbation of HF. The Cox proportional hazard model was used for statistical analyses. RESULTS: Univariate analysis showed that serum total bilirubin (TB, p < 0.01), serum B-type natriuretic peptide (p < 0.05), ejection fraction (EF, p < 0.05), and heart rate (p < 0.05) were significant predictors of hospital readmission secondary to acute decompensated HF. Multivariate analysis showed that high serum TB (> 1.3 mg/dL) on admission was an independent predictor (p < 0.05) of hospital readmission secondary to HF. The 'at-risk' group-patients with serum TB > 1.3 mg/dL and/or EF < 35% on admission-had a readmission rate that was 87% ± 20% (p < 0.05) higher than those with neither criterion. CONCLUSIONS: In patients with acute decompensated HF, elevated serum TB on admission with or without low EF (< 35%) predicts a worse prognosis and early future readmission, secondary to HF.