Search

Detecting 3D objects from point clouds is a practical yet challenging task that has attracted increasing attention recently. In this paper, we propose a Label-Guided auxiliary training method for 3D object detection (LG3D), which serves as an auxiliary network to enhance the feature learning of existing 3D object detectors. Specifically, we propose two novel modules: a Label-Annotation-Inducer that maps annotations and point clouds in bounding boxes to task-specific representations and a Label-Knowledge-Mapper that assists the original features to obtain detection-critical representations. The proposed auxiliary network is discarded in inference and thus has no extra computational cost at test time. We conduct extensive experiments on both indoor and outdoor datasets to verify the effectiveness of our approach. For example, our proposed LG3D improves VoteNet by 2.5% and 3.1% mAP on the SUN RGB-D and ScanNetV2 datasets, respectively., Comment: Yaomin Huang and Xinmei Liu are with equal contribution. This paper has been accepted by ECCV 2022

This paper presents a three-phase two-level high-power-density inverter operating at 1-kHz fundamental frequency and 100-kHz switching frequency under dynamic loads. The inverter was previously designed and optimized to operate at 400-Hz and capable to deliver 50 kW of power. Since the next-generation aerospace applications will likely operate at a higher fundamental frequency, some experimental tests have been done to evaluate the inverter operating under such conditions. This paper uses a simple voltage control to regulate the inverter output voltages under dynamic loading conditions. The experimental results show that the inverter output voltages can be well regulated even under load transients. The results also indicate that the previously designed inverter can still be operated at a higher fundamental frequency.

Aluminium Electrolytic Capacitors (AEC) are indispensable in applications where high capacitance per unit volume and low cost are preferred. However, it is also one of the most failure-prone components in a power converter system. As AEC is an electrochemical device, which follows Arrhenius law, the most important life-deciding factor of an AEC is its operating temperature [1]. In this paper, a condition monitoring system for AECs is proposed based on the accurate estimation of equivalent series resistance (ESR) and operating temperature. Operating temperature is estimated using the thermal impedance model based on improved power loss calculation methodology. The proposed method computes power loss using the capacitor voltage and current in contrast to the existing methods [4] - [7] which uses capacitor current and ESR to estimate the power loss. Thereby, it avoids the issue of ESR variation with ageing and temperature which may give erroneous results with ageing in case of stored look-up tables for ESR. ESR estimation is based on the method proposed in [14]. As ESR is the most widely used degradation indicator, the ESR and operating temperature can be used to estimate the present health status of an AEC.