Video Scene Detection Using Transformer Encoding Linker Network (TELNet).

This paper introduces a transformer encoding linker network (TELNet) for automatically identifying scene boundaries in videos without prior knowledge of their structure. Videos consist of sequences of semantically related shots or chapters, and recognizing scene boundaries is crucial for various video processing tasks, including video summarization. TELNet utilizes a rolling window to scan through video shots, encoding their features extracted from a fine-tuned 3D CNN model (transformer encoder). By establishing links between video shots based on these encoded features (linker), TELNet efficiently identifies scene boundaries where consecutive shots lack links. TELNet was trained on multiple video scene detection datasets and demonstrated results comparable to other state-of-the-art models in standard settings. Notably, in cross-dataset evaluations, TELNet demonstrated significantly improved results (F-score). Furthermore, TELNet's computational complexity grows linearly with the number of shots, making it highly efficient in processing long videos.

Investigation the electroplating behavior of self formed CuMn barrier.

The electrical and material properties of Copper (Cu) mixed with [0-10 atomic% manganese (Mn)] and pure Cu films deposited on silicon oxide (SiO2)/silicon (Si) are explored. Cu electroplating on self formed CuMn barrier was investigated with different Mn content. The electrochemical deposition of the Cu thin film onto the electrode using CuMn barrier was investigated. Scanning electron microscopic (SEM) micrographs of copper electroplating on CuMn films were examined, and the copper nucleation behaviors changed with the Mn content. Since the electrochemical impedance spectroscopy (EIS) is widely recognized as a powerful tool for the investigation of electrochemical behaviors, the tool was also used to verify the phenomena during plating. It was found that the charge-trasfer impedance decrease with the rise in the Mn content below 5%, but increase with the rise in the Mn content higher than 5%. The result was corresponded to the surface energy, the surface morphology, the corrosion and the oxidation of the substrate.

Optimization of Extrusion Variables and Maleic Anhydride Content on Biopolymer Blends Based on Poly(hydroxybutyrate-co-hydroxyvalerate)/Poly(vinyl acetate) with Tapioca Starch.

Poly(3-hyroxybutyrate-co-3-hydroxyvalerate) (PHBV), poly(vinyl acetate) (PVAc), and tapioca starch are environment-friendly materials. The present study used these materials to produce biodegradable plastic pellets by melt extrusion. The tapioca starch content of composite formulations, the maleic anhydride content, and the screw speed of the extruder were chosen as variables for the extrusion process. A Box-Behnken response surface design was used to establish mathematical models to predict the relationship between the operating variables and the objective attributes (tensile strength, elongation at break, and water absorption) of the blends. Blend morphology was also assessed. The regression coefficients revealed that the extrusion parameters most significantly affecting extrudate responses were tapioca starch content and maleic anhydride content, both showing significant (p < 0.01) linear effects. The results of the analysis of variance found the models are in good agreement with experimental results as informed by high correlation coefficients (R² > 0.9), with no significant lack of fit. From the numerical analysis, optimized operating variables (20.13% tapioca starch content, 10.14% maleic anhydride content, and a screw speed of 41.3 rpm) produced a product with optimum values of 16.4 MPa tensile strength, 13.2% elongation at break, and 30.94% water absorption.