Semi-supervised video object segmentation (semi-VOS) is required for many visual applications. This task tracking class-agnostic objects from a given mask. Various approaches have been developed and achieved high accuracy in this field, but these previous models are hard to be utilized real-world applications due slow inference time tremendous complexity. To significantly speed up while reducin...

Though Two-stage video object detectors cannot perform detection in real time, the accuracy of them is normally higher than that one-stage detectors. One essence two-stage can easily use feature information from adjacent frames to augment key frame features. How extract and exploit temporal features stream for needs further exploration. CenterNet an anchor-free detector regress bounding boxes h...

Abstract Video2Flink is a distributed highly scalable video processing system for bounded (i.e., stored) or unbounded continuous) and real-time streams with the same efficiency. It shows how complicated tasks can be expressed executed as pipelined data flows on Apache Flink, an open-source stream platform. uses Kafka to facilitate machine-to-machine (m2m) communication between production that r...