Titlebook: Neural Information Processing; 21st International C Chu Kiong Loo,Keem Siah Yap,Kaizhu Huang Conference proceedings 2014 Springer Internati

首創(chuàng)精神

Image Retrieval Using a Novel Color Similarity Measurement and Neural Networksith much lower time complexity. Compared with Fixed Cardinality (FC), Block Difference of Inverse Probabilities (BDIP) and Normalized Moment of Inertia (Nmi), our approach makes 7% improvement and obtains a better ANMRR (Average Normalized Modified Retrieval Rank).

逃避現(xiàn)實(shí)

Calibrating Independent Component Analysis with Laplacian Reference for Real-Time EEG Artifact RemovA. A proposed cut-off threshold for ICA is demonstrated to produce reliable and sound processing when compared to a Laplacian reference system. A methodology for real-time processing that is simple and efficient is being suggested.

intolerance

Real Time Crowd Counting with Human Detection and Human Trackingr detector. The detector can detect human in every frame with a high accuracy. Then human tracking is used to track the detected people and remove duplicates in successive frames. Experiments on a real-world video show the proposed method can give an accurate estimation in real time.

ALE

anaphylaxis

Forecasting Crowd State in Video by an Improved Lattice Boltzmann Modelction force term, which represents the crowds purpose drive toward the exit, is proposed. The model input is optical flow velocity field. Less error in the velocity fields computing and the capability in forecasting the crowd state is obtained.

Obituary

Properties of Multiobjective Robust Controller Using Difference Signals and Multiple Competitive Assues as well as the change of control objective by means of switching multiple CAN2s. However, the controller have not been analyzed enough. This paper clarifies several properties of the controller by means of examining the control of linear plants.

beta-carotene

Explicate

Image Denoising with Rectified Linear Unitsapturing patterns in natural images, our model can achieve better performance and less time consumption than those using sigmoid units. A large number of experiments show that our approach can achieve the state-of-the-art performance.

分開

領(lǐng)巾

Automatic Image Annotation Exploiting Textual and Visual Saliencyeen the labels. Then semantic communities and latent textual saliency are detected; For visual saliency, we adopt a dual-layer BoW (DL-BoW) model integrated with the local features and salient regions of the image. Experiments on NUS-WIDE dataset demonstrate that the proposed method outperforms other state-of-the-art approaches.