Finger Gesture Spotting from Long Sequences Based on Multi-Stream Recurrent Neural Networks

Gesture spotting is an essential task for recognizing finger gestures used to control in-car touchless interfaces. Automated methods to achieve this task require to detect video segments where gestures are observed, to discard natural behaviors of users’ hands that may look as target gestures, and be able to work online. In this paper, we address these challenges with a recurrent neural architecture for online finger gesture spotting. We propose a multi-stream network merging hand and hand-location features, which help to discriminate target gestures from natural movements of the hand, since these may not happen in the same 3D spatial location. Our multi-stream recurrent neural network (RNN) recurrently learns semantic information, allowing to spot gestures online in long untrimmed video sequences. In order to validate our method, we collect a finger gesture dataset in an in-vehicle scenario of an autonomous car. 226 videos with more than 2100 continuous instances were captured with a depth sensor. On this dataset, our gesture spotting approach outperforms state-of-the-art methods with an improvement of about 10% and 15% of recall and precision, respectively. Furthermore, we demonstrated that by combining with an existing gesture classifier (a 3D Convolutional Neural Network), our proposal achieves better performance than previous hand gesture recognition methods.

Survey on Autoconfiguration Schemes in IPV6 Based Manets

The MANETs are spontaneously formed infrastructure-less networks with self-configuring nodes. The network is characterized with a dynamic nature as the member nodes are highly mobile resulting in network merging and partitioning. Each node in MANETs acts as routers to assist in communication. To start with self-configurations, the autoconfiguration schemes are used in these networks to assign unique addresses to the nodes and allow them to communicate with other nodes. Due to the dynamic nature of MANETs and the intervention of new technologies in next generation networks, the existing autoconfiguration schemes face many issues. The research focuses to resolve the pertaining issues in autoconfiguration schemes are discussed in this book chapter.

Optimization Analysis on Urban Transportation Network of Lanzhou

On the basis of analyzing the model of C space network in Lanzhou, the small world characteristics were found. Combine the main idea of box counting method and weighted network merging mechanism, the evolution mechanism was put forward based on the self-similarity of the network, then the urban transportation network of Lanzhou is optimized by this method and optimization results are obtained.