Large language models (LLMs) have shown considerable potential in supporting medical diagnosis. However, their effective integration into clinical workflows is hindered by physicians’ difficulties in perceiving and trusting LLM capabilities, which often results in miscalibrated trust. Existing model evaluations primarily emphasize standardized benchmarks and predefined tasks, offering limited insights into clinical reasoning practices. Moreover, research on human–AI collaboration has rarely examined physicians’ perceptions of LLMs’ clinical reasoning capability. In this work, we investigate how physicians perceive LLMs’ capabilities in the clinical reasoning process. We designed clinical cases, collected the corresponding analyses, and obtained evaluations from physicians (N=37) to quantitatively represent their perceived LLM diagnostic capabilities. By comparing the perceived evaluations with benchmark performance, our study highlights the aspects of clinical reasoning that physicians value and underscores the limitations of benchmark-based evaluation. We further discuss the implications of opportunities for enhancing trustworthy collaboration between physicians and LLMs in LLM-supported clinical reasoning.

This paper studies the problem of how to assess the quality of photographing viewpoints and how to choose good viewpoints for taking photographs of architectures. We achieve this by learning from photographs of world famous landmarks that are available on the Internet and their viewpoint quality ranked by online user annotation. Unlike previous efforts devoted to photo quality assessment which mainly rely on 2D image features, we show in this paper combining 2D image features extracted from images with 3D geometric features computed on the 3D models can result in more reliable evaluation of viewpoint quality. Specifically, we collect a set of photographs for each of 15 world famous architectures as well as their 3D models from the Internet. Viewpoint recovery for images is carried out through an image-model registration process, after which a newly proposed viewpoint clustering strategy is exploited to validate users' viewpoint preferences when photographing landmarks. Finally, we extract a number of 2D and 3D features for each image based on multiple visual and geometric cues and perform viewpoint recommendation by learning from both 2D and 3D features using a specifically designed SVM-2K multi-view learner, achieving superior performance over using solely 2D or 3D features. We show the effectiveness of the proposed approach through extensive experiments. The experiments also demonstrate that our system can be used to recommend viewpoints for rendering textured 3D models of buildings for the use of architectural design, in addition to viewpoint evaluation of photographs and recommendation of viewpoints for photographing architectures in practice.

Denial of service (DoS) attacks, and jamming in particular, present a significant threat to wireless networks because they are easy to mount and difficult to detect and prevent. We present and analyze a special type of DoS attack, called random packet destruction (RPD) that works by transmitting short periods of noise signals. The RPD DoS attack can effectively shut down a wireless network. Since the attacker does not need to pretend to be a legal user participating in the network, current anti-attack measures such as encryption, authentication and authorization cannot prevent these types of attacks. RPD DoS attacks are pervasive in nature and can potentially be launched against any wireless networks that are detectable. An attacker can launch RPD attacks against wireless networks used for mission critical systems to inflict serious damages on lives or properties. The paper presents for the first time, both theoretical analysis and performance simulations of WLANs when operating under RPD DoS attacks for a range of types of network traffic.

In 802.11 WLANs, active intrusion attacks on the MAC layer causes changes in the management frame distributions. This paper focuses on detecting intrusions by analyzing the management traffic patterns. Experimental results are presented that describe the patterns generated by two well-known active attacks on 802.11 WLANs: deauthentication denial-of service (DoS) and man-in-the-middle (MITM). By clustering the management frame bursts over a certain period of time, we observe that the active attacking traffic patterns can be classified through calculation of a cluster content value (CCV). Our results show that when any station in a WLAN experiences clustering in its management traffic distribution, the CCV can be used to detect and classify the attacks the station is experiencing