In this study, we tackle Generalized Category Discovery (GCD) via a Relational Retrieval perspective, explicitly coupling labeled and unlabeled data through bidirectional knowledge transfer. While existing methods treat these sources separately, missing valuable interaction opportunities, we propose Relational Pattern Consistency (RPC) that enables mutual enhancement. RPC employs One-vs-All classifiers for soft ID/OOD decomposition, then introduces two mechanisms: (i) for known-class preservation, we transfer semantic behavioral alignment; (ii) for category discovery, we leverage the insight that samples from the same category maintain invariant relationships with known-class prototypes, transforming unreliable pseudo-labeling into well-defined relational pattern matching. This bidirectional design allows labeled data to guide unlabeled learning while discovering novel categories through their collective relational signatures. Extensive experiments demonstrate RPC achieves state-of-the-art performance on both generic and fine-grained benchmarks.

Knowledge distillation (KD), which distills knowledge from complex networks (teacher) to lightweight (student) networks, has been actively studied recently. Despite previous studies have proposed several advanced KD losses or intricate training strategies, the core concept of KD proves ineffective if the student model is too weak to mimic the teacher's performance. In this study, we aim to narrow the performance discrepancy between Transformer-based teacher and student models by incorporating the inductive biases of several heterogeneous student models. To this end, we put forward a novel cross-architecture knowledge distillation approach called Adaptive Cross-architecture Mutual Knowledge Distillation (ACMKD), which tries to mitigate the performance gap issue using a multi-students mutual learning strategy. Specifically, we utilize three mainstream models associated with various inductive biases (CNN, INN, and Transformer) as the student models. In addition, we propose an effective attention similarity mechanism to facilitate the student models in mimicking specific portions of the teacher model. Drawing inspiration from the Cannikin Law, we devise a unique second-stage KD process that dynamically enables the weakest student model to learn from other stronger student models again. We validate our proposed methods on ImageNet and CIFAR100 datasets, and the results confirm that our ACMKD method significantly narrows the performance gap compared to other KD methods.

Human activity recognition (HAR) based on multi-modal approach has been recently shown to improve the accuracy performance of HAR. However, restricted computational resources associated with wearable devices, i.e., smartwatch, failed to directly support such advanced methods. To tackle this issue, this study introduces an end-to-end Vision-to-Sensor Knowledge Distillation (VSKD) framework. In this VSKD framework, only time-series data, i.e., accelerometer data, is needed from wearable devices during the testing phase. Therefore, this framework will not only reduce the computational demands on wearable devices, but also produce a learning model that closely matches the performance of the computational expensive multi-modal approach. In order to retain the local temporal relationship and facilitate visual deep learning models, we first convert time-series data to two-dimensional images by applying the Gramian Angular Field (GAF) based encoding method. We adopted multi-scale TRN with BN-Inception and ResNet18 as the teacher and student network in this study, respectively. A novel loss function, named Distance and Angle-wised Semantic Knowledge loss (DASK), is proposed to mitigate the modality variations between the vision and the sensor domain. Extensive experimental results on UTD-MHAD, MMAct, and Berkeley-MHAD datasets demonstrate the competitiveness of the proposed VSKD model which can be deployed on wearable devices.

Wearable sensor-based Human Action Recognition (HAR) has achieved remarkable success recently. However, the accuracy performance of wearable sensor-based HAR is still far behind the ones from the visual modalities-based system (i.e., RGB video, skeleton and depth). Diverse input modalities can provide complementary cues and thus improve the accuracy performance of HAR, but how to take advantage of multi-modal data on wearable sensor-based HAR has rarely been explored. Currently, wearable devices, i.e., smartwatches, can only capture limited kinds of non-visual modality data. This hinders the multi-modal HAR association as it is unable to simultaneously use both visual and non-visual modality data. Another major challenge lies in how to efficiently utilize multi-modal data on wearable devices with their limited computation resources. In this work, we propose a novel Progressive Skeleton-to-sensor Knowledge Distillation (PSKD) model which utilizes only time-series data, i.e., accelerometer data, from a smartwatch for solving the wearable sensor-based HAR problem. Specifically, we construct multiple teacher models using data from both teacher (human skeleton sequence) and student (time-series accelerometer data) modalities. In addition, we propose an effective progressive learning scheme to eliminate the performance gap between teacher and student models. We also designed a novel loss function called Adaptive-Confidence Semantic (ACS), to allow the student model to adaptively select either one of the teacher models or the ground-truth label it needs to mimic. To demonstrate the effectiveness of our proposed PSKD method, we conduct extensive experiments on Berkeley-MHAD, UTD-MHAD and MMAct datasets. The results confirm that the proposed PSKD method has competitive performance compared to the previous mono sensor-based HAR methods.

The Internet of Things (IoT) is a rapidly growing system of physical sensors and connected devices, enabling advanced information gathering, interpretation, and monitoring. The realization of a versatile IoT edge computing framework will accelerate seamless integration of the cyber-world with new physical IoT devices, and will fundamentally change and empower the way humans interact with the world. While there are many cloud-based IoT computing frameworks, they cannot support the needs of IoT applications that require local processing and guarantee of consumer’s privacy. This article presents experimentation with the opensource plug-and-play IoT middleware, called Cordova Accesor Host. We demonstrated that Cordova Accessor Host supports the essential ingredients of the composition and reusability of IoT services using the accessor as the basic building block and adopting an accessor-module-plugin design pattern. The portability is demonstrated by using the same accessor for collecting sensor data from radically different IoT devices such as, wearables (e.g., smartwatches) and microcontrollers (e.g., Arduino). Our energy profiling experiments show that IoT services deployed using the Cordova Accessor Host consume around 35% less battery power than the same IoT services deployed in the native Android operating system.

Air pollution is a major global risk to human health and environment. Particle matter (PM) with diameters less than 2.5 micrometers (PM2.5) is more harmful to human health than other air pollutants because it can penetrate deeply into lungs and damage human respiratory system. A new imagebased deep feature analysis method is presented in this paper for PM2.5concentration estimation. Firstly, low level and high level features are extracted from images and their spectrums by a deep learning neural network, and then regression models are created using the extracted deep features to estimate the PM2.5concentrations, which are future refined by the collected weather information. The proposed method was evaluated using a PM2.5dataset with 1460 photos and the experimental results demonstrated that our method outperformed other state-of-the-art methods.