Search

With the explosive growth of video data in real applications, near-duplicate video retrieval (NDVR) has become indispensable and challenging, especially for short videos. However, all existing NDVR datasets are introduced for long videos. Furthermore, most of them are small-scale and lack of diversity due to the high cost of collecting and labeling near-duplicate videos. In this paper, we introduce a large-scale short video dataset, called SVD, for the NDVR task. SVD contains over 500,000 short videos and over 30,000 labeled videos of near-duplicates. We use multiple video mining techniques to construct positive/negative pairs. Furthermore, we design temporal and spatial transformations to mimic user-attack behavior in real applications for constructing more difficult variants of SVD. Experiments show that existing state-of-the-art NDVR methods, including real-value based and hashing based methods, fail to achieve satisfactory performance on this challenging dataset. The release of SVD dataset will foster research and system engineering in the NDVR area. The SVD dataset is available at https://svdbase.github.io.

Hashing has been widely used for large-scale search due to its low storage cost and fast query speed. By using supervised information, supervised hashing can significantly outperform unsupervised hashing. Recently, discrete supervised hashing and feature learning based deep hashing are two representative progresses in supervised hashing. On one hand, hashing is essentially a discrete optimization problem. Hence, utilizing supervised information to directly guide discrete (binary) coding procedure can avoid sub-optimal solution and improve the accuracy. On the other hand, feature learning based deep hashing, which integrates deep feature learning and hash-code learning into an end-to-end architecture, can enhance the feedback between feature learning and hash-code learning. The key in discrete supervised hashing is to adopt supervised information to directly guide the discrete coding procedure in hashing. The key in deep hashing is to adopt the supervised information to directly guide the deep feature learning procedure. However, most deep supervised hashing methods cannot use the supervised information to directly guide both discrete (binary) coding procedure and deep feature learning procedure in the same framework. In this paper, we propose a novel deep hashing method, called deep discrete supervised hashing (DDSH). DDSH is the first deep hashing method which can utilize pairwise supervised information to directly guide both discrete coding procedure and deep feature learning procedure and thus enhance the feedback between these two important procedures. Experiments on four real datasets show that DDSH can outperform other state-of-the-art baselines, including both discrete hashing and deep hashing baselines, for image retrieval.

Due to its low storage cost and fast query speed, cross-modal hashing (CMH) has been widely used for similarity search in multimedia retrieval applications. However, almost all existing CMH methods are based on hand-crafted features which might not be optimally compatible with the hash-code learning procedure. As a result, existing CMH methods with handcrafted features may not achieve satisfactory performance. In this paper, we propose a novel cross-modal hashing method, called deep crossmodal hashing (DCMH), by integrating feature learning and hash-code learning into the same framework. DCMH is an end-to-end learning framework with deep neural networks, one for each modality, to perform feature learning from scratch. Experiments on two real datasets with text-image modalities show that DCMH can outperform other baselines to achieve the state-of-the-art performance in cross-modal retrieval applications., 12 pages

Focal adhesion kinase (FAK), which plays a pivotal role in mediating cell proliferation, survival and migration, is frequently overexpressed in human colon cancer. In the present study, we utilized the short hairpin RNA (shRNA) to knock down the expression of FAK in SW480 human colon cancer cells in vitro. Furthermore, nude mice bearing human colon carcinoma SW480 were established and treated with plasmids encoding FAK shRNA encapsulated in DOTAP: Chol cationic liposome through tail vein injection. Tumor growth and potential side effect were observed during the treatment. Assessments of angiogenesis, cell proliferation, and apoptosis were performed by using immunohistochemistry against CD31, proliferating cell nuclear antigen (PCNA) and TUNEL assays, respectively. The results indicated that DOTAP: Chol could efficiently deliver the therapeutic plasmids systemically to tumor xenografts, resulting in suppression of tumor growth. Treatment with plasmid encoding FAK-targeted shRNA reduced mean tumor volume by approximately 86% compared with control groups (p < 0.01), accompanied with angiogenesis inhibition (p < 0.05), tumor cell proliferation suppression (p < 0.05) and apoptosis induction (p < 0.05). Taken together, our data demonstrated that shRNA-mediated silencing of FAK might be a potential therapeutic approach against human colon carcinoma.

The back-attack phenomenon of the gas side blowing streams and its influence on the erosion and wear of the refractory lining during the combined side and top blowing AOD refining process of stainless steel were investigated on a water model unit of a 120 t AOD converter. Sufficiently full kinetic similarity between the model and its prototype was maintained. The influences of the gas flow rates for side and top blowing, the side tuyere position and number were examined. The preliminary industrial experiments were conducted. The results indicated that the presence of a gas top blowing jet and using of multiple tuyeres did not change the basic features of the back-attack, but could give it some distinctive behaviors. The back-attack could indeed bring about the evident and uneven erosion and wear of the lining. On the back-attack, the gas streams of the main tuyeres had a decisive role, and the subtuyere streams showed a certain suppression and alleviation effect. At a given tuyere number and position, its frequency and pressure and the total average pulse number in per unit time in the processes of this work increased with an increase in the gas side blowing rate. At a given tuyere position and gas side blowing rate, the back-attack and its influence on the erosion and wear of the lining enhanced with decreasing the tuyere number (increasing the gas flow rate for single tuyere). The gas top blowing jet could make the back-attack become more uniform and its frequency decrease, and its intensity and the total average pulse number increase; and it could reduce the eroded and worn rate of the lining at a given tuyere number and position and gas side blowing rate. The increased amplitude of the back-attack intensity and the extension of the damaged area of the lining caused by the buoyancy in a combined blowing were smaller and lower than those in a simple side blowing. As another important reason resulted in the back-attack and the erosion and wear of the lining, the effectiveness of the circulatory motion of the liquid in a combined blowing is different from that in a simple side blowing. At a given tuyere number, properly increasing the angle between each tuyere could be beneficial to alleviating the back-attack and to slowing down the erosion and wear of the lining. Under the conditions of this work, the back-attack actions and the lining eroded and worn extents and rates with 7 tuyeres and 22.5° or 6 tuyeres and 27° were all gentler and lower than those with the other tuyere equipments and arrangements. The results obtained from the physical modeling studies on the refining process were reliable, believable and valid. Suitably increasing the angle between each tuyere of the 120 t AOD converter could raise the life of its lining by a big margin, and remarkably improve the technical and economic indications of the process.

To screen and obtain the potential human placenta antigens for the further application with serological analysis of recombinant cDNA expression library (SEREX).SEREX technique with some modifications was applied. In brief, immune sera from rabbit immunized with human chorionic tissue were used to screen human placenta tissue cDNA expression library. Positive clone plaques were obtained after two rounds of screen. And the positive clone plaques were purified and sequenced. BLAST software was applied to comparison the obtained sequences with those found in GenBank for bioinformatic analysis.69 positive clones were obtained from two rounds of screen. They were derived from 12 different genes, 9 of these were genes of known biology function, and 3 were genes of unknown biology function depending on the sequence analysis. Among the positive clones, chorionic somatomammotropin hormone 1 and 2 genes (CSH1 and CSH2) were found in 57 of positive clones (82.6%). This implied that the CSH1 and CSH2 might be the gene of encoding the important antigen and other genes obtained were related to the development of embryo.Modified xenogeneic immune SEREX technology is a very effective method to screen and isolate human placenta antigens. These antigens identified from this study might contribute to clarify the process of embryo development.