Automatic building extraction from high-resolution aerial and satellite images has many practical applications, such as urban planning and disaster management. However, the complex appearance and various scales of buildings in remote-sensing images bring a challenge for building extraction. In this study, we developed a novel multiscale building extraction method based on refined attention pyramid networks (RAPNets). We built an encoder–decoder structure, and combine atrous convolution, deformable convolution, attention mechanism, and pyramid pooling module to improve the performance of feature extraction in the encoding path. Moreover, the salient multiscale features were extracted by embedding the convolutional block attention module into the lateral connections. Finally, the refined feature pyramid structure was adopted in the decoding path to fuse the multiscale features to obtain the final extraction results. Experiments on two standard data sets (Inria aerial image labeling data set and xBD data set) show that our method achieves reliable results and outperforms the comparing methods.

The emerging technology of deep neural networks has been proven to be successful for hyperspectral image analysis. However, it is still a great challenge to apply the deep learning method for quantitatively retrieving mineralogical composition, because typical deep neural networks generally require thousands of labeled samples for training, while only a few mineral samples can be acquired and examined for quantitative examination in practice. To address this challenge, this letter proposes a training data augmentation approach which incorporates the prior-knowledge of hyperspectral reflectance characteristics using the classic Hapke equations. Experiments over both laboratory and airborne hyperspectral remote sensing data show that the proposed method outperforms the widely used approaches for quantitative mineral analysis.

We examined whether an other-race effect (ORE) exists in recognizing own groups and other ethnic groups emotional faces and whether such an effect depended on one's individual experience level. We used the learning-recognition paradigm and recruited 80 Tibetan students for each experiment to recognize the emotional faces of Tibetan and Han nationalities. Different types of ORE were observed when individuals recognize different emotional faces. On the one hand, positive emotional faces resulted in typical ORE, which was not influenced by an individual experience level. On the other hand, negative emotional faces resulted in typical ORE for low-experience individuals but resulted in a reverse ORE for high-experience individuals. Therefore, the recognized and identified emotional faces may differ in terms of how they are processed. In addition, the recognition of patterns from these images may be influenced by one's individual experience level.

This paper presents a generic error model, which is based on the product of exponentials (POEs) formula, for serial-robot calibration. The identifiability of parameters in this error model was analyzed. The analysis shows the following: 1) Errors in all joint twists are identifiable. 2) The joint zero-position errors and the initial transformation errors cannot be identified when they are involved in the same error model. With either or neither of them, three practicable error models were obtained. The joint zero-position errors are identifiable when the following condition is satisfied: Coordinates of joint twists are linearly independent. 3) The maximum number of identifiable parameters is 6n+ 6 for ann-degree-of-freedom (DOF) generic serial robot. Simulation results show the following: 1) The maximum number of identifiable parameters is 6r+ 3t+ 6, whereris the number of revolute joints, andtis the number of prismatic joints. 2) All the kinematic parameters of the selective compliant assembly robot arm (SCARA) robot and programmable universal machine for assembly (PUMA) 560 robots were identified by using the three error models, respectively. The error model based on the POE formula can be a complete, minimal, and continuous kinematic model for serial-robot calibration.