Search

Scene text recognition has been a hot research topic in computer vision due to its various applications. The state of the art is the attention-based encoder-decoder framework that learns the mapping between input images and output sequences in a purely data-driven way. However, we observe that existing attention-based methods perform poorly on complicated and/or low-quality images. One major reason is that existing methods cannot get accurate alignments between feature areas and targets for such images. We call this phenomenon "attention drift". To tackle this problem, in this paper we propose the FAN (the abbreviation of Focusing Attention Network) method that employs a focusing attention mechanism to automatically draw back the drifted attention. FAN consists of two major components: an attention network (AN) that is responsible for recognizing character targets as in the existing methods, and a focusing network (FN) that is responsible for adjusting attention by evaluating whether AN pays attention properly on the target areas in the images. Furthermore, different from the existing methods, we adopt a ResNet-based network to enrich deep representations of scene text images. Extensive experiments on various benchmarks, including the IIIT5k, SVT and ICDAR datasets, show that the FAN method substantially outperforms the existing methods., Comment: Revise the description of IC15 datasets (1811 samples)

Person re-identification is known as matching an individual captured in one or more cameras using a gallery of provided candidates from a different camera view. It is a hard task owing to variations in illumination, viewpoints, poses and small number of annotated training individuals. For obtaining the proper distance metrics, we propose a novel approach based on dictionary learning. Our method decomposes the dictionary into a view-invariant dictionary and a view-specific one in order to overcome the limited performance resulting from the balance between the discrimination and cross-view invariant ability. We present this multi-task joint dictionary learning method and show the competitive performance by comparing our results with state-of-the-art methods on two publicly available datasets.

In order to obtain a high photovoltaic (PV) efficiency, a solar cell must operate at both a high current and voltage. The current is determined by the semiconductor's ability to convert above-bandgap photons into electron-hole pairs that can be collected, while the maximum achievable voltage depends on maximizing the carrier densities and minimizing recombination within the cell. For a high quality semiconductor like GaAs, which has been shown to have an internal fluorescence yield of 99.7%, non-radiative recombination can be minimized to the point where the PV efficiency is limited by radiative emission from the cell. Here we show an improvement in output voltage and efficiency by engineering the spontaneous emission rate using photonic crystal structures. The proposed device is composed of a GaAs PV cell that has been nano-patterned with photonic crystals in order to control carrier spontaneous emission and, as a result, increase device output voltage. In the proposed device, this emission control is achieved by tuning the bandgap of the photonic crystal structure near the semiconductor band edge. Under these operating conditions, the open circuit voltage is increased by a factor of -(kBT/q) ln[FP], where FP is defined as the ratio of the spontaneous emission rate in the nano-patterned solar cell to the spontaneous emission rate in bulk GaAs. By engineering small FP the voltage of the device can be significantly improved, leading to photovoltaic efficiencies of ∼36% from a single junction device.