Jia-ji JIANG Hai-bin WAN Hong-min SUN Tuan-fa QIN Zheng-qiang WANG
In this paper, the Towards High Performance Voxel-based 3D Object Detection (Voxel-RCNN) three-dimensional (3D) point cloud object detection model is used as the benchmark network. Aiming at the problems existing in the current mainstream 3D point cloud voxelization methods, such as the backbone and the lack of feature expression ability under the bird’s-eye view (BEV), a high-performance voxel-based 3D object detection network (Reinforced Voxel-RCNN) is proposed. Firstly, a 3D feature extraction module based on the integration of inverted residual convolutional network and weight normalization is designed on the 3D backbone. This module can not only well retain more point cloud feature information, enhance the information interaction between convolutional layers, but also improve the feature extraction ability of the backbone network. Secondly, a spatial feature-semantic fusion module based on spatial and channel attention is proposed from a BEV perspective. The mixed use of channel features and semantic features further improves the network’s ability to express point cloud features. In the comparison of experimental results on the public dataset KITTI, the experimental results of this paper are better than many voxel-based methods. Compared with the baseline network, the 3D average accuracy and BEV average accuracy on the three categories of Car, Cyclist, and Pedestrians are improved. Among them, in the 3D average accuracy, the improvement rate of Car category is 0.23%, Cyclist is 0.78%, and Pedestrians is 2.08%. In the context of BEV average accuracy, enhancements are observed: 0.32% for the Car category, 0.99% for Cyclist, and 2.38% for Pedestrians. The findings demonstrate that the algorithm enhancement introduced in this study effectively enhances the accuracy of target category detection.
Zhishuo ZHANG Chengxiang TAN Xueyan ZHAO Min YANG
Entity alignment (EA) is a crucial task for integrating cross-lingual and cross-domain knowledge graphs (KGs), which aims to discover entities referring to the same real-world object from different KGs. Most existing embedding-based methods generate aligning entity representation by mining the relevance of triple elements, paying little attention to triple indivisibility and entity role diversity. In this paper, a novel framework named TTEA - Type-enhanced Ensemble Triple Representation via Triple-aware Attention for Cross-lingual Entity Alignment is proposed to overcome the above shortcomings from the perspective of ensemble triple representation considering triple specificity and diversity features of entity role. Specifically, the ensemble triple representation is derived by regarding relation as information carrier between semantic and type spaces, and hence the noise influence during spatial transformation and information propagation can be smoothly controlled via specificity-aware triple attention. Moreover, the role diversity of triple elements is modeled via triple-aware entity enhancement in TTEA for EA-oriented entity representation. Extensive experiments on three real-world cross-lingual datasets demonstrate that our framework makes comparative results.
Radar emitter identification (REI) is a crucial function of electronic radar warfare support systems. The challenge emphasizes identifying and locating unique transmitters, avoiding potential threats, and preparing countermeasures. Due to the remarkable effectiveness of deep learning (DL) in uncovering latent features within data and performing classifications, deep neural networks (DNNs) have seen widespread application in radar emitter identification (REI). In many real-world scenarios, obtaining a large number of annotated radar transmitter samples for training identification models is essential yet challenging. Given the issues of insufficient labeled datasets and abundant unlabeled training datasets, we propose a novel REI method based on a semi-supervised learning (SSL) framework with virtual adversarial training (VAT). Specifically, two objective functions are designed to extract the semantic features of radar signals: computing cross-entropy loss for labeled samples and virtual adversarial training loss for all samples. Additionally, a pseudo-labeling approach is employed for unlabeled samples. The proposed VAT-based SS-REI method is evaluated on a radar dataset. Simulation results indicate that the proposed VAT-based SS-REI method outperforms the latest SS-REI method in recognition performance.
Data sparsity has always been a problem in document classification, for which semi-supervised learning and few-shot learning are studied. An even more extreme scenario is to classify documents without any annotated data, but using only category names. In this paper, we introduce a nearest neighbor search-based method Con2Class to tackle this tough task. We intend to produce embeddings for predefined categories and predict category embeddings for all the unlabeled documents in a unified embedding space, such that categories can be easily assigned by searching the nearest predefined category in the embedding space. To achieve this, we propose confidence-driven contrastive learning, in which prompt-based templates are designed and MLM-maintained contrastive loss is newly proposed to finetune a pretrained language model for embedding production. To deal with the issue that no annotated data is available to validate the classification model, we introduce confidence factor to estimate the classification ability by evaluating the prediction confidence. The language model having the highest confidence factor is used to produce embeddings for similarity evaluation. Pseudo labels are then assigned by searching the semantically closest category name, which are further used to train a separate classifier following a progressive self-training strategy for final prediction. Our experiments on five representative datasets demonstrate the superiority of our proposed method over the existing approaches.
Qi QI Liuyi MENG Ming XU Bing BAI
In face super-resolution reconstruction, the interference caused by the texture and color of the hair region on the details and contours of the face region can negatively affect the reconstruction results. This paper proposes a semantic-based, dual-branch face super-resolution algorithm to address the issue of varying reconstruction complexities and mutual interference among different pixel semantics in face images. The algorithm clusters pixel semantic data to create a hierarchical representation, distinguishing between facial pixel regions and hair pixel regions. Subsequently, independent image enhancement is applied to these distinct pixel regions to mitigate their interference, resulting in a vivid, super-resolution face image.
2D and 3D semantic segmentation play important roles in robotic scene understanding. However, current 3D semantic segmentation heavily relies on 3D point clouds, which are susceptible to factors such as point cloud noise, sparsity, estimation and reconstruction errors, and data imbalance. In this paper, a novel approach is proposed to enhance 3D semantic segmentation by incorporating 2D semantic segmentation from RGB-D sequences. Firstly, the RGB-D pairs are consistently segmented into 2D semantic maps using the tracking pipeline of Simultaneous Localization and Mapping (SLAM). This process effectively propagates object labels from full scans to corresponding labels in partial views with high probability. Subsequently, a novel Semantic Projection (SP) block is introduced, which integrates features extracted from localized 2D fragments across different camera viewpoints into their corresponding 3D semantic features. Lastly, the 3D semantic segmentation network utilizes a combination of 2D-3D fusion features to facilitate a merged semantic segmentation process for both 2D and 3D. Extensive experiments conducted on public datasets demonstrate the effective performance of the proposed 2D-assisted 3D semantic segmentation method.
Haijun ZHOU Weixiang LI Ming CHENG Yuan SUN
Traditional intuitionistic fuzzy sets and hesitant fuzzy sets will lose some information while representing vague information, to avoid this problem, this paper constructs weighted generalized hesitant fuzzy sets by remaining multiple intuitionistic fuzzy values and giving them corresponding weights. For weighted generalized hesitant fuzzy elements in weighted generalized hesitant fuzzy sets, the paper defines some basic operations and proves their operation properties. On this basis, the paper gives the comparison rules of weighted generalized hesitant fuzzy elements and presents two kinds of aggregation operators. As for weighted generalized hesitant fuzzy preference relation, this paper proposes its definition and computing method of its corresponding consistency index. Furthermore, the paper designs an ensemble learning algorithm based on weighted generalized hesitant fuzzy sets, carries out experiments on 6 datasets in UCI database and compares with various classification algorithms. The experiments show that the ensemble learning algorithm based on weighted generalized hesitant fuzzy sets has better performance in all indicators.
Li HE Jingxuan ZHAO Jianyong DUAN Hao WANG Xin LI
In Natural Language Understanding, intent detection and slot filling have been widely used to understand user queries. However, current methods tend to rely on single words and sentences to understand complex semantic concepts, and can only consider local information within the sentence. Therefore, they usually cannot capture long-distance dependencies well and are prone to problems where complex intentions in sentences are difficult to recognize. In order to solve the problem of long-distance dependency of the model, this paper uses ConceptNet as an external knowledge source and introduces its extensive semantic information into the multi-intent detection and slot filling model. Specifically, for a certain sentence, based on confidence scores and semantic relationships, the most relevant conceptual knowledge is selected to equip the sentence, and a concept context map with rich information is constructed. Then, the multi-head graph attention mechanism is used to strengthen context correlation and improve the semantic understanding ability of the model. The experimental results indicate that the model has significantly improved performance compared to other models on the MixATIS and MixSNIPS multi-intent datasets.
Ren TAKEUCHI Rikima MITSUHASHI Masakatsu NISHIGAKI Tetsushi OHKI
The war between cyber attackers and security analysts is gradually intensifying. Owing to the ease of obtaining and creating support tools, recent malware continues to diversify into variants and new species. This increases the burden on security analysts and hinders quick analysis. Identifying malware families is crucial for efficiently analyzing diversified malware; thus, numerous low-cost, general-purpose, deep-learning-based classification techniques have been proposed in recent years. Among these methods, malware images that represent binary features as images are often used. However, no models or architectures specific to malware classification have been proposed in previous studies. Herein, we conduct a detailed analysis of the behavior and structure of malware and focus on PE sections that capture the unique characteristics of malware. First, we validate the features of each PE section that can distinguish malware families. Then, we identify PE sections that contain adequate features to classify families. Further, we propose an ensemble learning-based classification method that combines features of highly discriminative PE sections to improve classification accuracy. The validation of two datasets confirms that the proposed method improves accuracy over the baseline, thereby emphasizing its importance.
Yang YU Longlong LIU Ye ZHU Shixin CEN Yang LI
Pedestrian attribute recognition (PAR) aims to recognize a series of a person's semantic attributes, e.g., age, gender, which plays an important role in video surveillance. This paper proposes a multi-correlation graph convolutional network named MCGCN for PAR, which includes a semantic graph, visual graph, and synthesis graph. We construct a semantic graph by using attribute features with semantic constraints. A graph convolution is employed, based on prior knowledge of the dataset, to learn the semantic correlation. 2D features are projected onto visual graph nodes and each node corresponds to the feature region of each attribute group. Graph convolution is then utilized to learn regional correlation. The visual graph nodes are connected to the semantic graph nodes to form a synthesis graph. In the synthesis graph, regional and semantic correlation are embedded into each other through inter-graph edges, to guide each other's learning and to update the visual and semantic graph, thereby constructing semantic and regional correlation. On this basis, we use a better loss weighting strategy, the suit_polyloss, to address the imbalance of pedestrian attribute datasets. Experiments on three benchmark datasets show that the proposed approach achieves superior recognition performance compared to existing technologies, and achieves state-of-the-art performance.
Xuanke JIANG Sherief HASHIMA Kohei HATANO Eiji TAKIMOTO
In this paper, we investigate an online job scheduling problem with n jobs and k servers, where the accessibilities between the jobs and the servers are given as a bipartite graph. The scheduler is tasked with minimizing the regret, defined as the difference between the total flow time of the scheduler over T rounds and that of the best-fixed scheduling in hindsight. We propose an algorithm whose regret bounds are $O(n^2 sqrt{Tln (nk)})$ for general bipartite graphs, $O((n^2/k^{1/2}) sqrt{Tln (nk)})$ for the complete bipartite graphs, and $O((n^2/k) sqrt{T ln (nk)}$ for the disjoint star graphs, respectively. We also give a lower regret bound of $Omega((n^2/k) sqrt{T})$ for the disjoint star graphs, implying that our regret bounds are almost optimal.
Representation learning is a crucial and complex task for multivariate time series data analysis, with a wide range of applications including trend analysis, time series data search, and forecasting. In practice, unsupervised learning is strongly preferred owing to sparse labeling. However, most existing studies focus on the representation of individual subseries without considering relationships between different subseries. In certain scenarios, this may lead to downstream task failures. Here, an unsupervised representation learning model is proposed for multivariate time series that considers the semantic relationship among subseries of time series. Specifically, the covariance calculated by the Gaussian process (GP) is introduced to the self-attention mechanism, capturing relationship features of the subseries. Additionally, a novel unsupervised method is designed to learn the representation of multivariate time series. To address the challenges of variable lengths of input subseries, a temporal pyramid pooling (TPP) method is applied to construct input vectors with equal length. The experimental results show that our model has substantial advantages compared with other representation learning models. We conducted experiments on the proposed algorithm and baseline algorithms in two downstream tasks: classification and retrieval. In classification task, the proposed model demonstrated the best performance on seven of ten datasets, achieving an average accuracy of 76%. In retrieval task, the proposed algorithm achieved the best performance under different datasets and hidden sizes. The result of ablation study also demonstrates significance of semantic relationship in multivariate time series representation learning.
Ryota HIGASHIMOTO Soh YOSHIDA Takashi HORIHATA Mitsuji MUNEYASU
Noisy labels in training data can significantly harm the performance of deep neural networks (DNNs). Recent research on learning with noisy labels uses a property of DNNs called the memorization effect to divide the training data into a set of data with reliable labels and a set of data with unreliable labels. Methods introducing semi-supervised learning strategies discard the unreliable labels and assign pseudo-labels generated from the confident predictions of the model. So far, this semi-supervised strategy has yielded the best results in this field. However, we observe that even when models are trained on balanced data, the distribution of the pseudo-labels can still exhibit an imbalance that is driven by data similarity. Additionally, a data bias is seen that originates from the division of the training data using the semi-supervised method. If we address both types of bias that arise from pseudo-labels, we can avoid the decrease in generalization performance caused by biased noisy pseudo-labels. We propose a learning method with noisy labels that introduces unbiased pseudo-labeling based on causal inference. The proposed method achieves significant accuracy gains in experiments at high noise rates on the standard benchmarks CIFAR-10 and CIFAR-100.
Takefumi KAWAKAMI Takanori IDE Kunihito HOKI Masakazu MURAMATSU
In this paper, we apply two methods in machine learning, dropout and semi-supervised learning, to a recently proposed method called CSQ-SDL which uses deep neural networks for evaluating shift quality from time-series measurement data. When developing a new Automatic Transmission (AT), calibration takes place where many parameters of the AT are adjusted to realize pleasant driving experience in all situations that occur on all roads around the world. Calibration requires an expert to visually assess the shift quality from the time-series measurement data of the experiments each time the parameters are changed, which is iterative and time-consuming. The CSQ-SDL was developed to shorten time consumed by the visual assessment, and its effectiveness depends on acquiring a sufficient number of data points. In practice, however, data amounts are often insufficient. The methods proposed here can handle such cases. For the cases wherein only a small number of labeled data points is available, we propose a method that uses dropout. For those cases wherein the number of labeled data points is small but the number of unlabeled data is sufficient, we propose a method that uses semi-supervised learning. Experiments show that while the former gives moderate improvement, the latter offers a significant performance improvement.
Wataru KOBAYASHI Shigeru KANAZAWA Takahiko SHINDO Manabu MITSUHARA Fumito NAKAJIMA
We evaluated the energy efficiency per 1-bit transmission of an optical light source on InP substrate to achieve optical interconnection. A semiconductor optical amplifier (SOA) assisted extended reach EADFB laser (AXEL) was utilized as the optical light source to enhance the energy efficiency compared to the conventional electro-absorption modulator integrated with a DFB laser (EML). The AXEL has frequency bandwidth extendibility for operation of over 100Gbit/s, which is difficult when using a vertical cavity surface emitting laser (VCSEL) without an equalizer. By designing the AXEL for low power consumption, we were able to achieve 64-Gbit/s, 1.0pJ/bit and 128-Gbit/s, 1.5pJ/bit operation at 50°C with the transmitter dispersion and eye closure quaternary of 1.1dB.
Recent years have seen a decline in the art of analog IC design even though analog interface and analog signal processing remain just as essential as ever. While there are many contributing factors, four specific pressures which contribute the most to the loss of creativity and innovation within analog practice are examined: process evolution, risk aversion, digitally assisted analog, and corporate culture. Despite the potency of these forces, none are found to be insurmountable obstacles to reinvigorating the industry. A more creative future is within our reach.
Takuya WADATSUMI Kohei KAWAI Rikuu HASEGAWA Kikuo MURAMATSU Hiromu HASEGAWA Takuya SAWADA Takahito FUKUSHIMA Hisashi KONDO Takuji MIKI Makoto NAGATA
This paper presents on-chip characterization of electrostatic discharge (ESD) impacts applied on the Si-substrate backside of a flip-chip mounted integrated circuit (FC-IC) chip. An FC-IC chip has an open backside and there is a threat of reliability problems and malfunctions caused by the backside ESD. We prepared a test FC-IC chip and measured Si-substrate voltage fluctuations on its frontside by an on-chip monitor (OCM) circuit. The voltage surges as large as 200mV were observed on the frontside when a 200-V ESD gun was irradiated through a 5kΩ contact resistor on the backside of a 350μm thick Si substrate. The distribution of voltage heights was experimentally measured at 20 on-chip locations among thinned Si substrates up to 40μm, and also explained in full-system level simulation of backside ESD impacts with the equivalent models of ESD-gun operation and FC-IC chip assembly.
Kaito TOMARI Jun YONEDA Tetsuo KODERA
Reducing on-chip microwave crosstalk is crucial for semiconductor spin qubit integration. Toward crosstalk reduction and qubit integration, we investigate on-chip microwave crosstalk for gate electrode pad designs with (i) etched trenches between contact pads or (ii) contact pads with reduced sizes. We conclude that the design with feature (ii) is advantageous for high-density integration of semiconductor qubits with small crosstalk (below -25 dB at 6 GHz), favoring the introduction of flip-chip bonding.
Tadayoshi ENOMOTO Nobuaki KOBAYASHI
We developed a self-controllable voltage level (SVL) circuit and applied this circuit to a single-power-supply, six-transistor complementary metal-oxide-semiconductor static random-access memory (SRAM) to not only improve both write and read performances but also to achieve low standby power and data retention (holding) capability. The SVL circuit comprises only three MOSFETs (i.e., pull-up, pull-down and bypass MOSFETs). The SVL circuit is able to adaptively generate both optimal memory cell voltages and word line voltages depending on which mode of operation (i.e., write, read or hold operation) was used. The write margin (VWM) and read margin (VRM) of the developed (dvlp) SRAM at a supply voltage (VDD) of 1V were 0.470 and 0.1923V, respectively. These values were 1.309 and 2.093 times VWM and VRM of the conventional (conv) SRAM, respectively. At a large threshold voltage (Vt) variability (=+6σ), the minimum power supply voltage (VMin) for the write operation of the conv SRAM was 0.37V, whereas it decreased to 0.22V for the dvlp SRAM. VMin for the read operation of the conv SRAM was 1.05V when the Vt variability (=-6σ) was large, but the dvlp SRAM lowered it to 0.41V. These results show that the SVL circuit expands the operating voltage range for both write and read operations to lower voltages. The dvlp SRAM reduces the standby power consumption (PST) while retaining data. The measured PST of the 2k-bit, 90-nm dvlp SRAM was only 0.957µW at VDD=1.0V, which was 9.46% of PST of the conv SRAM (10.12µW). The Si area overhead of the SVL circuits was only 1.383% of the dvlp SRAM.
Daiki HIRATA Norikazu TAKAHASHI
Convolutional Neural Networks (CNNs) have shown remarkable performance in image recognition tasks. In this letter, we propose a new CNN model called the EnsNet which is composed of one base CNN and multiple Fully Connected SubNetworks (FCSNs). In this model, the set of feature maps generated by the last convolutional layer in the base CNN is divided along channels into disjoint subsets, and these subsets are assigned to the FCSNs. Each of the FCSNs is trained independent of others so that it can predict the class label of each feature map in the subset assigned to it. The output of the overall model is determined by majority vote of the base CNN and the FCSNs. Experimental results using the MNIST, Fashion-MNIST and CIFAR-10 datasets show that the proposed approach further improves the performance of CNNs. In particular, an EnsNet achieves a state-of-the-art error rate of 0.16% on MNIST.