Video Classification Using Deep Autoencoder Network

Abstract

We present a deep learning framework for video classification applicable to face recognition and dynamic texture recognition. A Deep Autoencoder Network Template (DANT) is designed whose weights are initialized by conducting unsupervised pre-training in a layer-wise fashion using Gaussian Restricted Boltzmann Machines. In order to obtain a class specific network and fine tune the weights for each class, the pre-initialized DANT is trained for each class of video sequences, separately. A majority voting technique based on the reconstruction error is employed for the classification task. The extensive evaluation and comparisons with state-of-theart approaches on Honda/UCSD, DynTex, and YUPPEN databases demonstrate that the proposed method significantly improves the performance of dynamic texture classification.

Design and Research of a Vehicle Approaching Reminder Device Based on ZigBee Wireless Sensor Network

“…e wireless sensor network expands people's ability to collect data, connects the physical data of the objective world to the transmission network, and provides humans with the most direct, efficient and reliable data on the next-generation network. is section is mainly a basic introduction to wireless sensor networks, mainly introducing the architecture, key features, design, and applications of wireless sensor networks [15]. e sensor node is usually an embedded system.…”