EXplainable AI (XAI) methods have been proposed to interpret how a deep neural network predicts inputs through model saliency explanations that highlight the input parts deemed important to arrive at a decision for a specific target. However, it remains challenging to quantify the correctness of their interpretability as current evaluation approaches either require subjective input from humans or incur high computation cost with automated evaluation. In this paper, we propose backdoor trigger patterns--hidden malicious functionalities that cause misclassification--to automate the evaluation of saliency explanations. Our key observation is that triggers provide ground truth for inputs to evaluate whether the regions identified by an XAI method are truly relevant to its output. Since backdoor triggers are the most important features that cause deliberate misclassification, a robust XAI method should reveal their presence at inference time. We introduce three complementary metrics for the systematic evaluation of explanations that an XAI method generates. We evaluate seven state-of-the-art model-free and model-specific post-hoc methods through 36 models trojaned with specifically crafted triggers using color, shape, texture, location, and size. We found six methods that use local explanation and feature relevance fail to completely highlight trigger regions, and only a model-free approach can uncover the entire trigger region. We made our code available at https://github.com/yslin013/evalxai.

Various domain users are increasingly leveraging real-time social media data to gain rapid situational awareness. However, due to the high noise in the deluge of data, effectively determining semantically relevant information can be difficult, further complicated by the changing definition of relevancy by each end user for different events. The majority of existing methods for short text relevance classification fail to incorporate users' knowledge into the classification process. Existing methods that incorporate interactive user feedback focus on historical datasets. Therefore, classifiers cannot be interactively retrained for specific events or user-dependent needs in real-time. This limits real-time situational awareness, as streaming data that is incorrectly classified cannot be corrected immediately, permitting the possibility for important incoming data to be incorrectly classified as well. We present a novel interactive learning framework to improve the classification process in which the user iteratively corrects the relevancy of tweets in real-time to train the classification model on-the-fly for immediate predictive improvements. We computationally evaluate our classification model adapted to learn at interactive rates. Our results show that our approach outperforms state-of-the-art machine learning models. In addition, we integrate our framework with the extended Social Media Analytics and Reporting Toolkit (SMART) 2.0 system, allowing the use of our interactive learning framework within a visual analytics system tailored for real-time situational awareness. To demonstrate our framework's effectiveness, we provide domain expert feedback from first responders who used the extended SMART 2.0 system.

In wireless networks, the cooperative relaying that provides a spatial diversity and improves the performance of system is a tendency in the future communications. There are three strategies in the cooperative communications. One is Amplify-and-Forward (AF) mode. Another is Decode-and-Forward (DF) mode and the other is Compress and Forward (CF) mode. Comparing with these three kind modes, AF mode is with the low complexity to implement. Consider the Raleigh fading environment where there are one information source, M relay stations and N destinations. The downlink signal includes two parts. One is from the source and the other comes from the relay stations. In order to achieve the maximum throughput in the system, it should obtain the maximum mutual information between the source and the destination. Hence, how to select an optimal relay station is important in the cooperative communications. This project proposes a relay selection algorithm with Amplify-and-Forward mode employed. The optimal selection scheme is based on the exhaustive search method. It is easy to be realized with a high computational complexity. In order to reduce the computational complexity, the relay station selection scheme with a maximum mutual information finding is proposed. With different relay station selection schemes, the performances are given for comparison.

Abstract A QR code is a two‐dimensional barcode that encodes information. A standard QR code contains only regular black and white squares, and thus is unattractive. This paper proposes a novel framework for embellishing a standard QR code, to make it both attractive and recognizable by any human while maintaining its scanability. The proposed method is inspired by artistic methods. A QR code is typically embellished by stylizing the squares and embedding images into it. In the proposed framework, the regular squares are reshaped using a binary examplar, to make their local appearances resemble the example shape. Additionally, an error‐aware warping technique for deforming the embedded image is proposed; it minimizes the error in the QR code that is generated by the embedding of the image to optimize the readability of the code. The proposed algorithm yields lower data error than previous global transformation techniques because the warping can locally deform the embedded image to conform to the squares that surround it. The proposed framework was examined by using it to embellish an extensive set of QR codes and to test the readability with various commercial QR code readers.

The portable electronic nose system is composed of a sensor array (sensor head) part and an electronics part. The sensors are made of polymer/mesoporous carbon composite materials for high sensitivity and selectivity. The electronics are sensor interface circuitry together with a microprocessor.