arXiv subjects

Weizhi Lu

Publications and source records attributed to Weizhi Lu.

9 recordsLinked to original sources

arXiv2025

Binary and Ternary Quantization Can Enhance Feature Discrimination

Quantization is widely applied in machine learning to reduce computational and storage costs for both data and models. Considering that classification tasks are fundamental to the field, it is crucial to investigate how quantization impacts classification performance. Traditional research has focused on quantization errors, assuming that larger errors generally lead to lower classification accuracy. However, this assumption lacks a solid theoretical foundation and often contradicts empirical observations. For example, despite introducing significant errors, $\{0,1\}$-binary and $\{0, \pm1\}$-ternary quantized data have sometimes achieved classification accuracy comparable or even superior to full-precision data. To reasonably explain this phenomenon, a more accurate evaluation of classification performance is required. To achieve this, we propose a direct analysis of the feature discrimination of quantized data, instead of focusing on quantization errors. Our analysis reveals that both binary and ternary quantization can potentially enhance, rather than degrade, the feature discrimination of the original data. This finding is supported by classification experiments conducted on both synthetic and real data.

cs.LG↗

arXiv2022

Ternary and Binary Quantization for Improved Classification

Dimension reduction and data quantization are two important methods for reducing data complexity. In the paper, we study the methodology of first reducing data dimension by random projection and then quantizing the projections to ternary or binary codes, which has been widely applied in classification. Usually, the quantization will seriously degrade the accuracy of classification due to high quantization errors. Interestingly, however, we observe that the quantization could provide comparable and often superior accuracy, as the data to be quantized are sparse features generated with common filters. Furthermore, this quantization property could be maintained in the random projections of sparse features, if both the features and random projection matrices are sufficiently sparse. By conducting extensive experiments, we validate and analyze this intriguing property.

cs.CV↗

arXiv2021

Cascaded Compressed Sensing Networks: A Reversible Architecture for Layerwise Learning

Recently, the method that learns networks layer by layer has attracted increasing interest for its ease of analysis. For the method, the main challenge lies in deriving an optimization target for each layer by inversely propagating the global target of the network. The propagation problem is ill posed, due to involving the inversion of nonlinear activations from lowdimensional to high-dimensional spaces. To address the problem, the existing solution is to learn an auxiliary network to specially propagate the target. However, the network lacks stability, and moreover, it results in higher complexity for network learning. In the letter, we show that target propagation could be achieved by modeling the network s each layer with compressed sensing, without the need of auxiliary networks. Experiments show that the proposed method could achieve better performance than the auxiliary network-based method.

cs.LG↗

arXiv2021

Deep Learning to Ternary Hash Codes by Continuation

Recently, it has been observed that {0,1,-1}-ternary codes which are simply generated from deep features by hard thresholding, tend to outperform {-1,1}-binary codes in image retrieval. To obtain better ternary codes, we for the first time propose to jointly learn the features with the codes by appending a smoothed function to the networks. During training, the function could evolve into a non-smoothed ternary function by a continuation method. The method circumvents the difficulty of directly training discrete functions and reduces the quantization errors of ternary codes. Experiments show that the generated codes indeed could achieve higher retrieval accuracy.

cs.CV↗

arXiv2019

Learning Actions from Human Demonstration Video for Robotic Manipulation

Learning actions from human demonstration is an emerging trend for designing intelligent robotic systems, which can be referred as video to command. The performance of such approach highly relies on the quality of video captioning. However, the general video captioning methods focus more on the understanding of the full frame, lacking of consideration on the specific object of interests in robotic manipulations. We propose a novel deep model to learn actions from human demonstration video for robotic manipulation. It consists of two deep networks, grasp detection network (GNet) and video captioning network (CNet). GNet performs two functions: providing grasp solutions and extracting the local features for the object of interests in robotic manipulation. CNet outputs the captioning results by fusing the features of both full frames and local objects. Experimental results on UR5 robotic arm show that our method could produce more accurate command from video demonstration than state-of-the-art work, thereby leading to more robust grasping performance.

cs.CV↗

arXiv2014

Sparse Matrix-based Random Projection for Classification

As a typical dimensionality reduction technique, random projection can be simply implemented with linear projection, while maintaining the pairwise distances of high-dimensional data with high probability. Considering this technique is mainly exploited for the task of classification, this paper is developed to study the construction of random matrix from the viewpoint of feature selection, rather than of traditional distance preservation. This yields a somewhat surprising theoretical result, that is, the sparse random matrix with exactly one nonzero element per column, can present better feature selection performance than other more dense matrices, if the projection dimension is sufficiently large (namely, not much smaller than the number of feature elements); otherwise, it will perform comparably to others. For random projection, this theoretical result implies considerable improvement on both complexity and performance, which is widely confirmed with the classification experiments on both synthetic data and real data.

cs.LG↗

arXiv2014

Bipartite Graph based Construction of Compressed Sensing Matrices

This paper proposes an efficient method to construct the bipartite graph with as many edges as possible while without introducing the shortest cycles of length equal to 4. The binary matrix associated with the bipartite graph described above presents comparable and even better phase transitions than Gaussian random matrices.

cs.IT↗

arXiv2014

Compressed Sensing Matrices: Binary vs. Ternary

Binary matrix and ternary matrix are two types of popular sensing matrices in compressed sensing for their competitive performance and low computation. However, to the best of our knowledge, there seems no literature aiming at evaluating their performances if they hold the same sparisty, though it is of practical importance. Based on both RIP analysis and numerical simulations, this paper, for the first time, discloses that {0, 1} binary matrix holds better overall performance over {0, +1, -1} ternary matrix, if they share the same distribution on nonzero positions.

cs.IT↗

arXiv2013

Near-optimal Binary Compressed Sensing Matrix

Compressed sensing is a promising technique that attempts to faithfully recover sparse signal with as few linear and nonadaptive measurements as possible. Its performance is largely determined by the characteristic of sensing matrix. Recently several zero-one binary sensing matrices have been deterministically constructed for their relative low complexity and competitive performance. Considering the complexity of implementation, it is of great practical interest if one could further improve the sparsity of binary matrix without performance loss. Based on the study of restricted isometry property (RIP), this paper proposes the near-optimal binary sensing matrix, which guarantees nearly the best performance with as sparse distribution as possible. The proposed near-optimal binary matrix can be deterministically constructed with progressive edge-growth (PEG) algorithm. Its performance is confirmed with extensive simulations.

cs.IT↗