We introduce a new large-scale data set of video URLs with densely-sampled object bounding box annotations called YouTube-BoundingBoxes (YT-BB). The data set consists of approximately 380,000 video segments about 19s long, automatically selected to feature objects in natural settings without editing or post-processing, with a recording quality often akin to that of a hand-held cell phone camera. The objects represent a subset of the COCO [32] label set. All video segments were human-annotated with high-precision classification labels and bounding boxes at 1 frame per second. The use of a cascade of increasingly precise human annotations ensures a label accuracy above 95% for every class and tight bounding boxes. Finally, we train and evaluate well-known deep network architectures and report baseline figures for per-frame classification and localization to provide a point of comparison for future work. We also demonstrate how the temporal contiguity of video can potentially be used to improve such inferences. The data set can be found at https://research.google.com/youtube-bb. We hope the availability of such large curated corpus will spur new advances in video object detection and tracking.
The maximum tangential strain (MTSN) criterion has been modified to include the effects of T‐stress and stress intensity factors in conditions of both plane stress and plane strain. Further, both the T‐stress and Poisson's ratio affecting the crack propagation are also discussed according to the extended MTSN (EMTSN) criterion, which is a modified MTSN criterion. Finally, the generalized maximum tangential stress (GMTS) criterion and the EMTSN criterion are used to predict the test results obtained with central cracked Brazilian disc (CCBD) specimens. The results indicate that the T‐stress and Poisson's ratio have a remarkable influence on the mixed mode fracture resistance based on the EMTSN criterion. Theoretical values of both the EMTSN and the GMTS criteria are in very good agreement with the test results. Moreover, the EMTSN criterion provides a better prediction for pure mode II.
Complex propagation patterns of hydraulic fractures often play important roles in naturally fractured formations due to complex mechanisms. Therefore, understanding propagation patterns and the geometry of fractures is essential for hydraulic fracturing design. In this work, a seepage–stress–damage coupled model based on the finite pore pressure cohesive zone (PPCZ) method was developed to investigate hydraulic fracture propagation behavior in a naturally fractured reservoir. Compared with the traditional finite element method, the coupled model with global insertion cohesive elements realizes arbitrary propagation of fluid-driven fractures. Numerical simulations of multiple-cluster hydraulic fracturing were carried out to investigate the sensitivities of a multitude of parameters. The results reveal that stress interference from multiple-clusters is responsible for serious suppression and diversion of the fracture network. A lower stress difference benefits the fracture network and helps open natural fractures. By comparing the mechanism of fluid injection, the maximal fracture network can be achieved with various injection rates and viscosities at different fracturing stages. Cluster parameters, including the number of clusters and their spacing, were optimal, satisfying the requirement of creating a large fracture network. These results offer new insights into the propagation pattern of fluid driven fractures and should act as a guide for multiple-cluster hydraulic fracturing, which can help increase the hydraulic fracture volume in naturally fractured reservoirs.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.