Bee Detection and Tracking at Hive Entrance using YOLO11 and ByteTrack
研究使用YOLO11和ByteTrack算法解决蜂巢入口蜜蜂检测与跟踪问题，通过适度数据增强、逐步解冻骨干网络及优化跟踪参数提高系统准确性。
0 citationsRead paper
Selected work
Representative Papers
Towards a physics-informed multiscale digital twin for precision medicine in Alzheimer's disease
This study addresses the challenges of elucidating brain aging drivers and individual trajectory heterogeneity in precision medicine for Alzheimer’s disease by constructing a physics-informed, multiscale digital twin platform. Innovatively integrating fatigue dissipation principles from materials science to quantify brain system stability, the framework employs an open, modular federated architecture to support mechanistic hypothesis testing. The platform enables quantitative assessment of residual compensatory capacity and system stability, facilitating the formal verification, comparison, and refinement of alternative mechanistic hypotheses. Ultimately, this work establishes a novel computational paradigm for dissecting the complex mechanisms underlying brain aging, offering a robust tool for advancing personalized therapeutic strategies in neurodegenerative diseases.
0 citationsRead paper
Towards Model-based Run-time Cybersecurity: On Control-Flow Anomaly Detection, Attack Identification, and Hardware Monitoring
This work addresses the vulnerability of conventional control-flow-based runtime monitoring to evasion through adversarial obfuscation, which often leads to detection failure or misclassification. To overcome this limitation, the paper proposes a model-driven, hardware-software co-designed security architecture. At the software layer, suspicious behaviors are initially identified through control-flow anomaly detection combined with attack-tree reasoning; concurrently, the hardware layer independently performs fine-grained control-flow validation. This synergistic approach effectively thwarts adversarial camouflage. In a case study involving authentication services, the framework accurately reclassifies a code injection attack—previously misjudged as a low-risk configuration issue—as a high-confidence control-flow hijacking event, thereby substantially enhancing both detection robustness and diagnostic precision.
0 citationsRead paper
MINT: Dynamic-Precision CNN Inference with MSDF Digit-Serial Arithmetic on FPGA
This work addresses the computational redundancy and energy-efficiency limitations of conventional fixed-precision CNNs on FPGAs by proposing a dynamic-precision inference accelerator based on Most-Significant-Digit-First (MSDF) serial arithmetic. The design integrates redundant signed-digit representation with a budget-constrained greedy search algorithm to dynamically select the minimal feasible integer precision (ranging from INT2 to INT7) per layer and terminate computation early once the target accuracy is achieved. To the best of our knowledge, this is the first integration of MSDF arithmetic with dynamic-precision CNN inference, enabling on-demand precision control. Implemented on a Zynq-7020 FPGA, the approach achieves throughputs of 19.86 and 18.86 GOPS and energy efficiencies of 29.51 and 26.40 GOPS/W for VGG-16 and ResNet-18, respectively, using average precisions of 5.64 and 6.04 bits—yielding over 60% higher energy efficiency than INT8 baselines with less than 2% accuracy loss.
0 citationsRead paper
Cross-View Urban Traffic Dataset: Drone-Supervised Ground Truth for Monocular Bird's-Eye View Localization
This work addresses the cross-view perception challenges of identity alignment for urban traffic objects and monocular-to-bird’s-eye-view localization across street-level and aerial perspectives. The authors introduce a novel dataset comprising synchronized first-person bicycle videos and drone-captured aerial footage, offering the first identity-level aligned data across such extreme viewpoints. The pipeline leverages synchronized multi-view acquisition, trajectory-level annotations, and inverse perspective mapping, combined with MonoLayout-inspired learning and regression models to enable cross-view identity matching and bird’s-eye-view prediction from monocular images under aerial supervision. Experiments demonstrate high recall in cross-view matching, though performance is limited by over-allocation and temporal inconsistency; monocular prediction significantly improves with aerial supervision yet leaves room for optimization in lightweight settings. The accompanying standardized evaluation protocol, annotation toolkit, and baseline methods aim to advance research in cross-view urban traffic understanding.
0 citationsRead paper
Recent publications
Latest Papers
Bee Detection and Tracking at Hive Entrance using YOLO11 and ByteTrack
研究使用YOLO11和ByteTrack算法解决蜂巢入口蜜蜂检测与跟踪问题，通过适度数据增强、逐步解冻骨干网络及优化跟踪参数提高系统准确性。
0 citationsRead paper
Towards a physics-informed multiscale digital twin for precision medicine in Alzheimer's disease
This study addresses the challenges of elucidating brain aging drivers and individual trajectory heterogeneity in precision medicine for Alzheimer’s disease by constructing a physics-informed, multiscale digital twin platform. Innovatively integrating fatigue dissipation principles from materials science to quantify brain system stability, the framework employs an open, modular federated architecture to support mechanistic hypothesis testing. The platform enables quantitative assessment of residual compensatory capacity and system stability, facilitating the formal verification, comparison, and refinement of alternative mechanistic hypotheses. Ultimately, this work establishes a novel computational paradigm for dissecting the complex mechanisms underlying brain aging, offering a robust tool for advancing personalized therapeutic strategies in neurodegenerative diseases.
0 citationsRead paper
Towards Model-based Run-time Cybersecurity: On Control-Flow Anomaly Detection, Attack Identification, and Hardware Monitoring
This work addresses the vulnerability of conventional control-flow-based runtime monitoring to evasion through adversarial obfuscation, which often leads to detection failure or misclassification. To overcome this limitation, the paper proposes a model-driven, hardware-software co-designed security architecture. At the software layer, suspicious behaviors are initially identified through control-flow anomaly detection combined with attack-tree reasoning; concurrently, the hardware layer independently performs fine-grained control-flow validation. This synergistic approach effectively thwarts adversarial camouflage. In a case study involving authentication services, the framework accurately reclassifies a code injection attack—previously misjudged as a low-risk configuration issue—as a high-confidence control-flow hijacking event, thereby substantially enhancing both detection robustness and diagnostic precision.
0 citationsRead paper
MINT: Dynamic-Precision CNN Inference with MSDF Digit-Serial Arithmetic on FPGA
This work addresses the computational redundancy and energy-efficiency limitations of conventional fixed-precision CNNs on FPGAs by proposing a dynamic-precision inference accelerator based on Most-Significant-Digit-First (MSDF) serial arithmetic. The design integrates redundant signed-digit representation with a budget-constrained greedy search algorithm to dynamically select the minimal feasible integer precision (ranging from INT2 to INT7) per layer and terminate computation early once the target accuracy is achieved. To the best of our knowledge, this is the first integration of MSDF arithmetic with dynamic-precision CNN inference, enabling on-demand precision control. Implemented on a Zynq-7020 FPGA, the approach achieves throughputs of 19.86 and 18.86 GOPS and energy efficiencies of 29.51 and 26.40 GOPS/W for VGG-16 and ResNet-18, respectively, using average precisions of 5.64 and 6.04 bits—yielding over 60% higher energy efficiency than INT8 baselines with less than 2% accuracy loss.
0 citationsRead paper
Cross-View Urban Traffic Dataset: Drone-Supervised Ground Truth for Monocular Bird's-Eye View Localization
This work addresses the cross-view perception challenges of identity alignment for urban traffic objects and monocular-to-bird’s-eye-view localization across street-level and aerial perspectives. The authors introduce a novel dataset comprising synchronized first-person bicycle videos and drone-captured aerial footage, offering the first identity-level aligned data across such extreme viewpoints. The pipeline leverages synchronized multi-view acquisition, trajectory-level annotations, and inverse perspective mapping, combined with MonoLayout-inspired learning and regression models to enable cross-view identity matching and bird’s-eye-view prediction from monocular images under aerial supervision. Experiments demonstrate high recall in cross-view matching, though performance is limited by over-allocation and temporal inconsistency; monocular prediction significantly improves with aerial supervision yet leaves room for optimization in lightweight settings. The accompanying standardized evaluation protocol, annotation toolkit, and baseline methods aim to advance research in cross-view urban traffic understanding.
0 citationsRead paper