Libo Huang — Page 3

arXiv2021

Multi-Scale Cost Volumes Cascade Network for Stereo Matching

Stereo matching is essential for robot navigation. However, the accuracy of current widely used traditional methods is low, while methods based on CNN need expensive computational cost and running time. This is because different cost volumes play a crucial role in balancing speed and accuracy. Thus we propose MSCVNet, which combines traditional methods and neural networks to improve the quality of cost volume. Concretely, our network first generates multiple 3D cost volumes with different resolutions and then uses 2D convolutions to construct a novel cascade hourglass network for cost aggregation. Meanwhile, we design an algorithm to distinguish and calculate the loss for discontinuous areas of disparity result. According to the KITTI official website, our network is much faster than most top-performing methods (24 times than CSPN, 44 times than GANet, etc.). Meanwhile, compared to traditional methods (SPS-St, SGM) and other real-time stereo matching networks (Fast DS-CS, DispNetC, and RTSNet, etc.), our network achieves a big improvement in accuracy, demonstrating the feasibility and capability of the proposed method.

cs.CV↗

arXiv2020

A Unified Model of Feature Extraction and Clustering for Spike Sorting

Spike sorting plays an irreplaceable role in understanding brain codes. Traditional spike sorting technologies perform feature extraction and clustering separately after spikes are well detected. However, it may often cause many additional processes and further lead to low-accurate and/or unstable results especially when there are noises and/or overlapping spikes in datasets. To address these issues, in this paper, we proposed a unified optimisation model integrating feature extraction and clustering for spike sorting. Interestingly, instead of the widely used combination strategies, i.e., performing the principal component analysis (PCA) for spike feature extraction and K-means (KM) for clustering in sequence, we unified PCA and KM into one optimisation model, which reduces additional processes with fewer iteration times. Subsequently, by embedding the K-means++ strategy for initialising and a comparison updating rule in the solving process, the proposed model can well handle the noises and/or overlapping interference. Finally, taking the best of the clustering validity indices into the proposed model, we derive an automatic spike sorting method. Plenty of experimental results on both synthetic and real-world datasets confirm that our proposed method outperforms the related state-of-the-art approaches.

eess.SP↗