Artificial intelligence (AI) radar technology offers several advantages over other technologies, including low cost, privacy assurance, high accuracy, and environmental resilience. One challenge faced by AI radar technology is the high cost of equipment and the lack of radar datasets for deep-learning model training. Moreover, conventional radar signal processing methods have the obstacles of poor resolution or complex computation. Therefore, this paper discusses an innovative approach in the integration of radar technology and machine learning for effective surveillance systems that can surpass the aforementioned limitations. This approach is detailed into three steps: signal acquisition, signal processing, and feature-based classification. A hardware prototype of the signal acquisition circuitry was designed for a Continuous Wave (CW) K-24 GHz frequency band radar sensor. The collected radar motion data was categorized into non-human motion, human walking, and human walking without arm swing. Three signal processing techniques, namely short-time Fourier transform (STFT), mel spectrogram, and mel frequency cepstral coefficients (MFCCs), were employed. The latter two are typically used for audio processing, but in this study, they were proposed to obtain micro-Doppler spectrograms for all motion data. The obtained micro-Doppler spectrograms were then fed to a simplified 2D convolutional neural networks (CNNs) architecture for feature extraction and classification. Additionally, artificial neural networks (ANNs) and 1D CNN models were implemented for comparative analysis on various aspects. The experimental results demonstrated that the 2D CNN model trained on the MFCC feature outperformed the other two methods. The accuracy rate of the object classification models trained on micro-Doppler features was 97.93%, indicating the effectiveness of the proposed approach.

LiDAR panoptic segmentation (LPS) offers a comprehensive perception of the driving environment for autonomous vehicles, providing valuable and relevant information for downstream tasks such as scene understanding or 4D panoptic tracking. Current state-of-the-art (SOTA) methods are predominantly based on the detection-transformer (DETR) paradigm. These approaches formulate panoptic segmentation as a problem of predicting the association scores of points within each group, which can be either a thing class instance or a stuff class. To this end, groups are represented by a fixed number of randomly initialized vectors called queries, whose values remain constant during inference. In this approach, queries are required to represent a wide range of objects that present in any particular scene. Recent studies on DETR application in the image domain have attempted to resolve this issue by introducing a strategy for adaptive query generation on a per-scene basis, which has achieved promising results, while in the LiDAR domain, existing DETR-based approaches have yet to explore adaptive mechanisms for queries generation. Motivated by that insight, our work proposes a technique for query representation called clustered feature aggregation (CFA) - a method for computing the features associated with an instance and using them as a query for the segmentation of that instance. To cluster points belonging to the same group, this study introduces shifted point clustering (SPC), an enhancement of traditional clustering techniques, as it secures high-confidence instance accumulation. Our proposed methods, in combination with a strong DETRbased baseline, establish new SOTA results for the panoptic segmentation challenge, outperforming previous top-performers by 1.20% and 1.94%, in PQ metric, on SemanticKITTI and Panoptic Nuscenes datasets, respectively

LiDAR perception for autonomous driving applications offers highly accurate scene depiction in three-dimensional (3D) spaces, whose most representative task is LiDAR panoptic segmentation (LPS), as it offers exhibition of both instance- and semantic-level segmentation in a holistic manner. Although previous approaches have achieved mature performance, no research has explored temporal information for enhancing LPS performance. As multi-frame processing can assist in better predictions in terms of feature representation and recursive forecasting, which has been proven in other LiDAR perception challenges, this study proposes an effective and temporal-aware panoptic segmentation method for LiDAR point clouds. Specifically, we introduce two modules: convolution-based cross-frame fusion attention (CFFA) and adjacent shifted feature encoder (ASFE) modules. The CFFA module can fuse multi-frame features on the basis of the idea of convolution-based attention, whereas the ASFE module leverages adjacent model outputs and serves as an intermediate guide for final segmentation predictions. Consequent to our extensive experiments, the two modules have been reaffirmed in terms of their productivity in the realm of the LPS. The proposed LPS model achieves impressive panoptic-quality metric scores that are evaluated on different popular benchmarks (63.36% under SemanticKITTI and 78.54% under Panoptic nuScenes), outperforming previous state-of-the-art methods by a significant margin. Further quantitative and qualitative analyses provide evidence of the advantages of multi-frame processing for the LPS together with demonstrations of its particular behavior under different settings.