We introduce modifications to state-of-the-art approaches to aggregating local video descriptors by using attention mechanisms and function approximations. Rather than using ensembles of existing architectures, we provide an insight on creating new architectures. We demonstrate our solutions in the "The 2nd YouTube-8M Video Understanding Challenge", by using frame-level video and audio descriptors. We obtain testing accuracy similar to the state of the art, while meeting budget constraints, and touch upon strategies to improve the state of the art. Model implementations are available in https://github.com/pomonam/LearnablePoolingMethods.

No abstract

No abstract

No abstract

Video is playing a more and more important role in future Vehicular Ad-hoc Network (VANET) communication. It is a feasible medium for information sharing and entertainment (infotainment) with its high capacity, consistency and influence on human beings. This paper introduces the SVC coding scheme to VANET video streaming. We propose an Optimal Scheduling Algorithm (OSA) for video streaming in highway VANET scenarios using scalable video coding (SVC). We conduct extensive simulations to evaluate the performance of OSA. Simulation results show that OSA outperforms all the competing schemes over variant density and velocity highway VANET.