Large Language Models (LLMs) are increasingly used not only to generate code, but also to judge it: comparing, ranking, or scoring competing solutions.However, their reliability in this evaluative role remains poorly understood.Inconsistent or flawed judgments can undermine benchmarks and distort training signals.This paper investigates the performance and robustness of LLMs when used as code judges.We introduce CodeJudgeBench, a benchmark explicitly designed to evaluate LLM-as-a-Judge models across three critical coding tasks: code generation, code repair, and unit test generation.We comprehensively benchmark the performance of 26 LLM-as-a-Judge models, encompassing general-purpose, code-tuned, and reasoning models.Our empirical findings reveal that relatively small reasoning models (e.g., Qwen3-8B) can outperform much larger non-reasoning models up to 70B.We further stress-test robustness by applying both general and code-specific perturbations.All models show significant instability and are sensitive to changes such as response ordering, variable naming, and misleading comments.These findings highlight serious concerns about the consistency and robustness of LLM-based judges for coding tasks.

Semantic communication has emerged as a promising paradigm exhibiting improved robustness compared to traditional approaches under low SNR conditions. Precise synchronization is imperative for accurate semantic communication. However, existing synchronization techniques face challenges reliably achieving synchronization at low SNRs, limiting semantic communication development. To improve synchronization performance, especially under low SNR scenarios, this work proposes an image-based semantic synchronization method (SeSy) leveraging inherent image correlations. SeSy is applicable to both semantic and traditional communication systems. Theoretical analysis establishes bounds on the miss detected ratio (MDR) for SeSy. Experimental results demonstrate that SeSy achieves lower MDR and root mean square error (RMSE) compared to traditional methods across various SNR levels, especially at low SNRs.

Multiple access technologies are critical technologies in every communication era. As a promising paradigm for next-generation mobile communication, semantic communication has explored new semantic information space resources. Based on the characteristic that different semantic models cannot understand semantic information generated by other models, we propose the concept of semantic orthogonal signals. Combining the advantages of Deep joint source and channel coding (DeepJSCC), an Orthogonal-Model Division Multiple Access (O-MDMA) technology that can be applied to any semantic model is proposed. The essence of O-MDMA is to migrate the anti-interference capability of DeepJSCC to the multi-user capacity. Compared with Non-Orthgonal Multiple Access (NOMA) and Model Division Multiple Access (MDMA) technologies, O-MDMA has better performance. The O-MDMA can be integrated with NOMA, and experimental results show that the combined technique can save more bandwidth.

Amid data privacy concerns, Federated Learning (FL) has emerged as a promising machine learning paradigm that enables privacy-preserving collaborative model training. However, there exists a need for a platform that matches data owners (supply) with model requesters (demand). In this paper, we present CrowdFL, a platform to facilitate the crowdsourcing of FL model training. It coordinates client selection, model training, and reputation management, which are essential steps for the FL crowdsourcing operations. By implementing model training on actual mobile devices, we demonstrate that the platform improves model performance and training efficiency. To the best of our knowledge, it is the first platform to support crowdsourcing-based FL on edge devices.

As countries enter the endemic phase of COVID-19, people's risk of exposure to the virus is greater than ever. There is a need to make more informed decisions in our daily lives on avoiding crowded places. Crowd monitoring systems typically require costly infrastructure. We propose a crowd-sourced crowd monitoring platform which leverages user inputs to generate crowd counts and forecast location crowdedness. A key challenge for crowd-sourcing is a lack of incentive for users to contribute. We propose a Reinforcement Learning based dynamic incentive mechanism to optimally allocate rewards to encourage user participation.

Alzheimer's disease (AD) is a chronic neurodegen-erative disease that causes cognitive deficits, which severely interfere with daily life. Convolutional Neural Networks (CNNs) have been used to analyze Medical Resonance Imaging (MRI) scans for the early detection of AD. Prior works have explored supervised pre-training, unsupervised pre-training, and joint training to improve the diagnostic accuracy of CNNs. However, there is no consensus on the best approach. We compare the different pre-training methods in a standardized setting. Our experiments find that supervised pre-training and joint training outperform unsupervised pre-training when data is extremely limited. With more data, unsupervised pre-training closes the performance gap and, in some cases, outperforms supervised pre-training and joint training. In addition, we propose a simple hybrid approach of unsupervised pre-training followed by joint training that achieves the best performance.

The COVID-19 pandemic is one of the most severe challenges the world faces today. In order to contain the transmission of COVID-19, people around the world have been advised to practise social distancing. However, maintaining social distance is a challenging problem, as we often do not know beforehand how crowded the places we intend to visit are. In this paper, we demonstrate crowded.sg, an AI-empowered platform that leverages on Unmanned Aerial Vehicles (UAVs), crowdsourced images, and computer vision techniques to provide social distancing decision support.

Dementia is one of the major causes of disability and dependency among older people. Early detection is the key for preserving the quality of life of the patients and reducing caring costs. The Clock Drawing Test (CDT) is commonly used by clinicians to screen for early signs of dementia. We build an automated CDT that runs on mobile platforms, enabling convenient and frequent self-monitoring and testing at minimal costs. Our system combines both a spatial-temporal approach and a purely image-based deep learning approach to analyze and evaluate the hand-drawn clocks based on established clinical criteria. Our system produces scores that are highly correlated with expert human raters.

Parkinson’s disease (PD) is a chronic disease with a high risk of incidence after the age of 60 and is a problem for many countries facing an aging population. Current works have mainly focused on supervised learning using data collected from various sensors to differentiate between PD and healthy subjects. However, such supervised methods are not ideal for prognosis where there are no labels (i.e., we do not know in advance which subjects will develop PD in the future). We propose to tackle the problem as a semi-supervised anomaly detection task, where we model the physiological patterns of healthy subjects instead. A self-supervised learning technique first learns a good representation of the sensor signals. The representations are then adapted to capture inter-class patterns for anomaly detection. Evaluation on a large-scale PD dataset shows that our approach can learn discriminative features.

The COVID-19 pandemic has disrupted the lives of millions across the globe. In Singapore, promoting safe distancing by managing crowds in public areas have been the cornerstone of containing the community spread of the virus. One of the most important solutions to maintain social distancing is to monitor the crowdedness of indoor and outdoor points of interest. Using Nanyang Technological University (NTU) as a testbed, we develop and deploy a platform that provides live and predicted crowd counts for key locations on campus to help users plan their trips in an informed manner, so as to mitigate the risk of community transmission.

Federated Edge Learning (FEL) is a distributed Machine Learning (ML) framework for collaborative training on edge devices. FEL improves data privacy over traditional centralized ML model training by keeping data on the devices and only sending local model updates to a central coordinator for aggregation. However, challenges still remain in existing FEL architectures where there is high communication overhead between edge devices and the coordinator. In this paper, we present a working prototype of blockchain-empowered and communication-efficient FEL framework, which enhances the security and scalability towards large-scale implementation of FEL.

Population aging is becoming an increasingly important issue around the world. As people live longer, they also tend to suffer from more challenging medical conditions. Currently, there is a lack of a holistic technology-powered solution for providing quality care at affordable cost to patients suffering from co-morbidity. In this paper, we demonstrate a novel AI-powered solution to provide early detection of the onset of Dementia + Parkinson's disease (DPD) co-morbidity, a condition which severely limits a senior's ability to live actively and independently. We investigate useful in-game behaviour markers which can support machine learning-based predictive analytics on seniors' risk of developing DPD co-morbidity.