The inherent sparsity of LiDAR data often leads to extremely sparse depth maps, which poses a challenge for the development of LiDAR-based egocentric vehicles, such as self-driving cars and mobile robots. To overcome ...
详细信息
The inherent sparsity of LiDAR data often leads to extremely sparse depth maps, which poses a challenge for the development of LiDAR-based egocentric vehicles, such as self-driving cars and mobile robots. To overcome this limitation, guided depth completion methods use calibrated camera images to create precise, dense depth predictions from sparse LiDAR data. However, the extreme reliance on camera image limits its generalization of guided depth completion, especially its robustness to weather and light. In this paper, we aim to utilize camera images only during training phase to improve unguided depth completion, and discard camera in the inference phase. We comprehensively analyze the pivotal role of camera images in the depth completion task and emphasize the significance of the frequency distribution within the local windows, quantitatively demonstrating its substantial contribution. Subsequently, we introduce cross-modality knowledge distillation to align LiDAR features with camera features in the frequency domain, yielding corresponding guidance features. We devise a guidance and selection module to mitigate unavoidable inaccuracies in knowledge distillation, while it can enhance depth features and adeptly selects more precise encoded values from both the guidance branch and the unguided input. To further refine the completion result, we propose a progressive depth completion module incorporating two sub-networks connected by an attention for refinement module. This module produces weighted features from the decoder of the first stage to enhance the features in the encoder of the second stage. We denominate our method as Better Unguided Network (BUNet) and evaluate its efficacy on the KITTI depth completion benchmark and NYUv2 dataset, demonstrating its superiority over methods that exclude camera images during the inference phase. IEEE
This article presents a novel adaptive full-order sliding-mode observer (AFSMO) tailored for sensorless control of permanent magnet synchronous motors (PMSMs). Additionally, an online compensation scheme for inverter ...
详细信息
A tree-based support generation algorithm based on Fused Deposition Modeling is proposed to optimize support material consumption. The algorithm first identifies dangerous facets and partitions them. Support points ar...
详细信息
In the sensorless permanent magnet synchronous motor (PMSM) control system, the additional noise caused by high-frequency (HF) injection method greatly affects the wide application in various fields. The noise reducti...
详细信息
Predictive control is widely used in dc-dc buck converters. With different control values, conventional voltage predictive control and average voltage predictive control are reviewed, respectively. The former has a fa...
详细信息
The debugging of high-speed trains is a process to ensure the stable operation of trains, and traditional debugging equipment has some problems in terms of comprehensiveness, reliability, and incomplete testing. In or...
详细信息
This article proposes a dual internal model decoupling control strategy to solve the high order and strong coupling problem of the permanent magnet synchronous motor (PMSM) system fed by the current source inverter (C...
详细信息
In order to find the shunt malfunction fault of track circuit in time and accurately, this paper proposes a prediction method of shunting malfunction of track circuit based on wavelet packet energy entropy and ELM. Fi...
详细信息
Investigating the methods of handling real-world objects with robotic arms is a crucial aspect of artificial intelligence research. By inputting images from cameras or other devices, neural networks extract suitable e...
详细信息
The hydraulic hose assembly consists of hoses, hose fittings and other accessories, and is an important component of the hydraulic system. With the wide application of hydraulic system in various fields, the quality a...
详细信息
暂无评论