Large Model Size Research Articles

Advancing Ton-Bag Detection in Seaport Logistics with an Enhanced YOLOv8 Algorithm

Intelligent logistics and freight transportation is an important part of realizing the intelligence of port terminals. Due to the problems of inaccurate ton bag identification, high costs, large model sizes, and long computation times in traditional freight transportation—issues that hinder meeting real-time requirements on resource-constrained operational equipment—this paper proposes an improved lightweight ton bag detection algorithm, YOLOv8-TB (YOLOv8-Ton Bag), which is optimized based on YOLOv8. Firstly, the improved LZKAC module is introduced to combine with SPPF to form a new SPPFLKZ module, which improves the feature expression performance. Then, with reference to spatial and channel reconstruction convolution and deformable convolution, the C2f-SCTT block is designed for the backbone network, which reduces the spatial and channel redundancy between features in the network. Finally, the C2f-ORECZ block based on a linear scaling layer is designed for the neck, which reduces the training overhead and strengthens the feature learning of the feature extraction network for the targets in the complex background of the harbor and adds the 160 × 160 scale detection head to strengthen small target detection abilities. On the logistics ton bag operation dataset provided by shipping port enterprises, the improved algorithm improves by 3.7% and 5% compared with the original algorithm in mAP50 and mAP50-95, respectively, the model size is reduced by 4.42 MB and the amount of model computation is only 8 G, which is capable of accurately detecting logistics ton bags in real time. The superiority of the method is verified by comparing it with other classical target detection algorithms.

A lightweight framework for unsupervised anomalous sound detection based on selective learning of time-frequency domain features

For industrial anomalous sound detection (ASD), self-supervised methods have achieved significant detection performance in many cases. Nevertheless, these methods typically rely on the availability of external auxiliary information, and they may not work when such information are not feasible. Unsupervised methods do not leverage auxiliary information, whereas they usually obtained lower detection performance compared to self-supervised ones. Though some unsupervised methods have shown potential performance improvements, they are at the cost of complex implementation or large model sizes. As to the issues, this paper presents an unsupervised ASD method based on spectrogram frames selection (SFS) and AutoEncoder for Frequency-feature Selection (AEFS), called SFS-AEFS. First, SFS is developed based upon the temporal characteristics of machine sounds, which aims to select spectrogram frames (SFs) that contains the primary sound information while discarding the portions that are affected by noises or interferences or do not contain the target sound. Next, AEFS is developed by introducing a Scaling Gate (SG) after AE. For the selected SF features, AEFS aims to selectively enhance the mode learning of partial frequency dimensions and weaken the rest ones. Comparative experiments with the current ASD methods were made on the DCASE 2020 Challenge Task2 dataset. The related results demonstrate that our method achieved the best performance among all relevant unsupervised methods and is comparable to the current SOTA self-supervised methods. Moreover, our method is lightweight with model parameters being only 0.08MB.

Improved Road Target Detection Algorithm for YOLOv7-Tiny

In complex road scenes, we propose an enhanced road target detection algorithm for YOLOv7-Tiny to address issues such as large model size, misclassification, and low localization accuracy. Our approach involves several key modifi-cations. Firstly, we replace the LeakyReLU activation function in YOLOv7-Tiny with H-swish. This replacement not only reduces the number of parameters in the model but also enhances its feature extraction capabilities. Additionally, we replace the ELAN module in the Neck with the DPCH-ELAN module and introduce the darknetblock module along with the pcconv convolution. These modifications improve the network's ability to comprehend complex patterns and semantics, thereby enabling it to capture features at different levels of input data. Moreover, we introduce the pcconv convolution building block at the output side to handle heterogeneous information in complex road scenes, thereby enhancing the network's performance in detecting road targets and abnormalities. In our experiments using a mul-ti-source dataset, the improved model exhibits a reduction in GFLOPs by 20.90% and a decrease in the number of pa-rameters by 24.49%. Furthermore, the mean average precision scores (map) at thresholds of 0.5 and 0.5~0.9 are im-proved to 77.3% and 51.8%, respectively, compared to the original YOLOv7-Tiny model. These experimental results demonstrate that our enhanced model achieves a reduction in model size while simultaneously enhancing detection accuracy, thereby meeting the requirements for real-time detection. To assess the generalizability of our approach, we conducted comparison experiments on the VOC2012 dataset. The results indicate that the improved algorithm exhibits robust generalization capabilities across different datasets.

Visual speech recognition using compact hypercomplex neural networks

Recent progress in visual speech recognition systems due to advances in deep learning and large-scale public datasets has led to impressive performance compared to human professionals. The potential applications of these systems in real-life scenarios are numerous and can greatly benefit the lives of many individuals. However, most of these systems are not designed with practicality in mind, requiring large-size models and powerful hardware, factors which limit their applicability in resource-constrained environments and other real-world tasks. In addition, few works focus on developing lightweight systems that can be deployed in such conditions. Considering these issues, we propose compact networks that take advantage of hypercomplex layers that utilize a sum of Kronecker products to reduce overall parameter demands and model sizes. We train and evaluate our proposed models on the largest public dataset for single word speech recognition for English. Our experiments show that high compression rates are achievable with a minimal accuracy drop, indicating the method’s potential for practical applications in lower-resource environments. Code and models are available at https://github.com/jpanagos/vsr_phm.

Improved Lightweight YOLOv5 Algorithm for Detecting Grid Inspection Object

Abstract In the field of unmanned aerial vehicle-based transmission line inspection, the neural network models currently employed suffer from large model sizes and high computational requirements, resulting in a trade-off between detection speed and accuracy. We propose an improved lightweight YOLOv5 object detection algorithm based on channel pruning to address this issue. The approach incorporates the SimAM attention mechanism into the backbone network to enhance feature extraction capability without increasing computational complexity, thereby improving detection accuracy. Traditional convolutional modules in the neck network are replaced with GSconv modules to reduce convolutional computation and enhance detection speed. The Wise-IoU loss is utilized as the localization loss for the detection model, speeding up model convergence. Finally, a model compression technique is implemented to eliminate unnecessary channels within the network, resulting in a decrease in model size and a boost in detection speed. Experimental results on a constructed grid inspection dataset demonstrate that the improved model achieves a 72.1% reduction in model size, a 62.5% reduction in GFLOPs, a 5.9 FPS improvement, and a 0.3% mAP increase compared to the original YOLOv5. These findings underscore the advantages of the improved lightweight YOLOv5 object detection algorithm, offering a blend of heightened accuracy, streamlined size, and swift detection capabilities.

Disruptive atomic jumps induce grain boundary stagnation

Grain growth in polycrystalline materials can be impeded by grain boundary (GB) stagnation. Using atomistic simulations, we observed that during GB migration, the disruptive jumps of a few GB atoms can disturb the original ordered collective movement of GB atoms, leading to the stagnation of the entire GB. These disruptive atomic jumps can be activated by both high driving forces and high temperatures, with even jumps of a few atoms capable of causing the stagnation of an entire GB. This mechanism also explains the non-Arrhenius behavior observed in some GBs. Additionally, a large model size could increase the rate of disruptive atomic jumps, and a clear transition in thermal behavior is observed with the increase of the GB size in GBs exhibiting clear thermally activated stagnation. Our further investigation shows that the disruptive atoms involved in these jumps do not differ from other GB atoms in terms of atomic energy, volume, density, local entropy, or Voronoi tessellation, and no “jam transition” was observed in the energy barrier spectra. This fact makes those disruptive jumps challenging to detect. To address this issue, we propose a displacement vector analysis method that effectively identifies these subtle disruptive jumps.

Photonics for Neuromorphic Computing: Fundamentals, Devices, and Opportunities.

In the dynamic landscape of Artificial Intelligence (AI), two notable phenomena are becoming predominant: the exponential growth of large AI model sizes and the explosion of massive amount of data. Meanwhile, scientific research such as quantum computing and protein synthesis increasingly demand higher computing capacities. As the Moore's Law approaches its terminus, there is an urgent need for alternative computing paradigms that satisfy this growing computing demand and break through the barrier of the von Neumann model. Neuromorphic computing, inspired by the mechanism and functionality of human brains, uses physical artificial neurons to do computations and is drawing widespread attention. This review studies the expansion of optoelectronic devices on photonic integration platforms that has led to significant growth in photonic computing, where photonic integrated circuits (PICs) have enabled ultrafast artificial neural networks (ANN) with sub-nanosecond latencies, low heat dissipation, and high parallelism. In particular, various technologies and devices employed in neuromorphic photonic AI accelerators, spanning from traditional optics to PCSEL lasers are examined. Lastly, it is recognized that existing neuromorphic technologies encounter obstacles in meeting the peta-level computing speed and energy efficiency threshold, and potential approaches in new devices, fabrication, materials, and integration to drive innovation are also explored. As the current challenges and barriers in cost, scalability, footprint, and computing capacity are resolved one-by-one, photonic neuromorphic systems are bound to co-exist with, if not replace, conventional electronic computers and transform the landscape of AI and scientific computing in the foreseeablefuture.

Polyp-LVT: Polyp segmentation with lightweight vision transformers

Automatic segmentation of polyps in endoscopic images is crucial for early diagnosis and surgical planning of colorectal cancer. However, polyps closely resemble surrounding mucosal tissue in both texture and indistinct borders and vary in size, appearance, and location which possess great challenge to polyp segmentation. Although some recent attempts have been made to apply Vision Transformer (ViT) to polyp segmentation and achieved promising performance, their application in clinical scenarios is still limited by high computational complexity, large model size, redundant dependencies, and significant training costs. To address these limitations, we propose a novel ViT-based approach named Polyp-LVT, strategically replacing the attention layer in the encoder with a global max pooling layer, which significantly reduces the model’s parameter count and computational cost while keeping the performance undegraded. Furthermore, we introduce a network block, named Inter-block Feature Fusion Module (IFFM), into the decoder, aiming to offer a streamlined yet highly efficient feature extraction. We conduct extensive experiments on three public polyp image benchmarks to evaluate our method. The experimental results show that compared with the baseline models, our Polyp-LVT network achieves a nearly 44% reduction in model parameters while gaining comparable segmentation performance.

Axial partial compressive behavior of round-ended concrete-filled steel tubular columns

The paper conducted 19 partially and 5 fully loaded tests on round-ended concrete-filled steel tubular (CFST) columns. These tests revealed that the ultimate strengths of round-ended CFST columns under partial load declined as the cross-sectional aspect ratio or partial compression area ratio increased. However, the use of a top plate effectively enhanced their ultimate strengths. In addition, a finite element model (FEM) was utilized to generate 487 large-size models to explore the confinement effect, stress distribution, and parametric effects. Since there are no specific design standards available for round-ended CFST columns, the paper has evaluated the applicability of design methods provided in Eurocode 4, AISC 360–16, and Chinese code GB50396 which are intended for circular, square, or elliptical CFST columns for round-ended CFST columns under partial load. However, these existing design techniques showed inaccurate and dispersed resistance predictions for round-ended CFST columns. Therefore, an effective simplified method for forecasting the resistance predictions of round-ended CFST columns under the partial load has been proposed based on the testing results and FEM analysis.

An Improved YOLOv5s Model for Building Detection

With the continuous advancement of autonomous vehicle technology, the recognition of buildings becomes increasingly crucial. It enables autonomous vehicles to better comprehend their surrounding environment, facilitating safer navigation and decision-making processes. Therefore, it is significant to improve detection efficiency on edge devices. However, building recognition faces problems such as severe occlusion and large size of detection models that cannot be deployed on edge devices. To solve these problems, a lightweight building recognition model based on YOLOv5s is proposed in this study. We first collected a building dataset from real scenes and the internet, and applied an improved GridMask data augmentation method to expand the dataset and reduce the impact of occlusion. To make the model lightweight, we pruned the model by the channel pruning method, which decreases the computational costs of the model. Furthermore, we used Mish as the activation function to help the model converge better in sparse training. Finally, comparing it to YOLOv5s (baseline), the experiments show that the improved model reduces the model size by 9.595 MB, and the mAP@0.5 reaches 82.3%. This study will offer insights into lightweight building detection, demonstrating its significance in environmental perception, monitoring, and detection, particularly in the field of autonomous driving.

Non-Autoregressive Line-Level Code Completion

Software developers frequently use code completion tools to accelerate software development by suggesting the following code elements. Researchers usually employ AutoRegressive (AR) decoders to complete code sequences in a left-to-right, token-by-token fashion. To improve the accuracy and efficiency of code completion, we argue that tokens within a code statement have the potential to be predicted concurrently. In this article, we first conduct an empirical study to analyze the dependency among the target tokens in line-level code completion. The results suggest that it is potentially practical to generate all statement tokens in parallel. To this end, we introduce SANAR, a simple and effective syntax-aware non-autoregressive model for line-level code completion. To further improve the quality of the generated code, we propose an adaptive and syntax-aware sampling strategy to boost the model’s performance. The experimental results obtained from two widely used datasets indicate that our model outperforms state-of-the-art code completion approaches of similar model size by a considerable margin, and is faster than these models with up to 9× speed-up. Moreover, the extensive results additionally demonstrate that the enhancements achieved by SANAR become even more pronounced with larger model sizes, highlighting their significance.

PCa-RadHop: A transparent and lightweight feed-forward method for clinically significant prostate cancer segmentation

Prostate Cancer is one of the most frequently occurring cancers in men, with a low survival rate if not early diagnosed. PI-RADS reading has a high false positive rate, thus increasing the diagnostic incurred costs and patient discomfort. Deep learning (DL) models achieve a high segmentation performance, although require a large model size and complexity. Also, DL models lack of feature interpretability and are perceived as “black-boxes” in the medical field. PCa-RadHop pipeline is proposed in this work, aiming to provide a more transparent feature extraction process using a linear model. It adopts the recently introduced Green Learning (GL) paradigm, which offers a small model size and low complexity. PCa-RadHop consists of two stages: Stage-1 extracts data-driven radiomics features from the bi-parametric Magnetic Resonance Imaging (bp-MRI) input and predicts an initial heatmap. To reduce the false positive rate, a subsequent stage-2 is introduced to refine the predictions by including more contextual information and radiomics features from each already detected Region of Interest (ROI). Experiments on the largest publicly available dataset, PI-CAI, show a competitive performance standing of the proposed method among other deep DL models, achieving an area under the curve (AUC) of 0.807 among a cohort of 1,000 patients. Moreover, PCa-RadHop maintains orders of magnitude smaller model size and complexity.

Integration of Semantic and Topological Structural Similarity Comparison for Entity Alignment without Pre-Training

Entity alignment (EA) is a critical task in integrating diverse knowledge graph (KG) data and plays a central role in data-driven AI applications. Traditional EA approaches rely on entity embeddings, but their effectiveness is limited by scarce KG input data and representation learning techniques. Large language models have shown promise, but face challenges such as high hardware requirements, large model sizes and computational inefficiency, which limit their applicability. To overcome these limitations, we propose an entity-alignment model that compares the similarity between entities by capturing both semantic and topological information to enable the alignment of entities with high similarity. First, we analyze descriptive information to quantify semantic similarity, including individual features such as types and attributes. Then, for topological analysis, we introduce four conditions based on graph connectivity and structural patterns to determine subgraph similarity within three hops of the entity’s neighborhood, thereby improving accuracy. Finally, we integrate semantic and topological similarity using a weighted approach that considers dataset features. Our model requires no pre-training and is designed to be compact and generalizable to different datasets. Experimental results on four standard EA datasets validate the effectiveness of our proposed model.

Calibrated confidence learning for large-scale real-time crash and severity prediction

Real-time crash and severity prediction is a complex task, and there is no existing framework to predict crash likelihood and severity together. Creating such a framework poses numerous challenges, particularly not independent and identically distributed (non-IID) data, large model sizes with high computational costs, missing data, sensitivity vs. false alarm rate (FAR) trade-offs, and real-world deployment strategies. This study introduces a novel modeling technique to address these challenges and develops a deployable real-world framework. We used extensive real-time traffic and weather data to develop a crash likelihood prediction modeling prototype, leveraging our preliminary work of spatial ensemble modeling. Next, we equipped this spatial ensemble model with local model regularization to calibrate model confidence training. The investigated regularizations include weight decay, label smoothing and knowledge distillation. Furthermore, post-calibration on model outputs was conducted to improve severity rating identification. We tested the framework to predict crashes and severity in real-time, categorizing crashes into four levels. Results were compared with benchmark models, real-world deployment mechanisms were explained, traffic safety improvement potential and sustainability aspects of the study were discussed. Modeling results demonstrated excellent performance, and fatal, severe, minor and PDO crash severities were predicted with 91.7%, 83.3%, 85.6%, and 87.7% sensitivity, respectively, and with very low FAR. Similarly, the viability of our model to predict different severity levels for specific crash types, i.e., all-crash types, rear-end crashes, and sideswipe/angle crashes, were examined, and it showed excellent performance. Our modeling technique showed great potential for reducing model size, lowering computational costs, improving sensitivity, and, most importantly, reducing FAR. Finally, the deployment strategy for the proposed crash and severity prediction technique is discussed, and its potential to predict crashes with severity levels in real-time will make a substantial contribution to tailoring specific strategies to prevent crashes.

Comparing penalization methods for linear models on large observational health data.

This study evaluates regularization variants in logistic regression (L1, L2, ElasticNet, Adaptive L1, Adaptive ElasticNet, Broken adaptive ridge [BAR], and Iterative hard thresholding [IHT]) for discrimination and calibration performance, focusing on both internal and external validation. We use data from 5 US claims and electronic health record databases and develop models for various outcomes in a major depressive disorder patient population. We externally validate all models in the other databases. We use a train-test split of 75%/25% and evaluate performance with discrimination and calibration. Statistical analysis for difference in performance uses Friedman's test and critical difference diagrams. Of the 840 models we develop, L1 and ElasticNet emerge as superior in both internal and external discrimination, with a notable AUC difference. BAR and IHT show the best internal calibration, without a clear external calibration leader. ElasticNet typically has larger model sizes than L1. Methods like IHT and BAR, while slightly less discriminative, significantly reduce model complexity. L1 and ElasticNet offer the best discriminative performance in logistic regression for healthcare predictions, maintaining robustness across validations. For simpler, more interpretable models, L0-based methods (IHT and BAR) are advantageous, providing greater parsimony and calibration with fewer features. This study aids in selecting suitable regularization techniques for healthcare prediction models, balancing performance, complexity, and interpretability.

Coded aperture compressive temporal imaging via unsupervised lightweight local-global networks with geometric characteristics

Coded aperture compressive temporal imaging (CACTI) utilizes compressive sensing (CS) theory to compress three dimensional (3D) signals into 2D measurements for sampling in a single snapshot measurement, which in turn acquires high-dimensional (HD) visual signals. To solve the problems of low quality and slow runtime often encountered in reconstruction, deep learning has become the mainstream for signal reconstruction and has shown superior performance. Currently, however, impressive networks are typically supervised networks with large-sized models and require vast training sets that can be difficult to obtain or expensive. This limits their application in real optical imaging systems. In this paper, we propose a lightweight reconstruction network that recovers HD signals only from compressed measurements with noise and design a block consisting of convolution to extract and fuse local and global features, stacking multiple features to form a lightweight architecture. In addition, we also obtain unsupervised loss functions based on the geometric characteristics of the signal to guarantee the powerful generalization capability of the network in order to approximate the reconstruction process of real optical systems. Experimental results show that our proposed network significantly reduces the model size and not only has high performance in recovering dynamic scenes, but the unsupervised video reconstruction network can approximate its supervised version in terms of reconstruction performance.

Simulation and reconstruction for 3D elastic wave using multi-GPU and CUDA-aware MPI

3D finite-difference time-domain numerical simulation and reconstruction based on the domain decomposition technique are essential parts of high-performance computation for reverse-time migration and full-waveform inversion. However, the low GPU utilization in computing for small-sized models and the tremendous memory consumption for large-sized models may result in low computational efficiency and high memory costs. This paper proposes a contiguous memory management (CMM) method and a variable-order wavefield reconstruction (VWR) method. The CMM allocates the memory of many small-sized arrays used for MPI communications on a larger-sized contiguous memory block, which aims to reduce the number of MPI communications between subdomains and improve the communication bandwidth, thus reducing the MPI time overhead and improving the GPU utilization. Meanwhile, the VWR can flexibly set the number of layers of boundary wavefield used for source wavefield reconstruction according to the host memory capacity and accuracy requirements. Since one layer of boundary wavefield could be stored using the VWR, the memory consumption of host memory can be significantly alleviated. Numerical experiments show that GPU utilization in computing for the model with a size of 1213 can be improved from 25% to 90% using the CMM method, and the VWR method can reduce memory consumption by about 86% while maintaining good accuracy in wavefield reconstruction. In addition, the issue of how to obtain a domain decomposition scheme with optimal performance is discussed in this paper.

Effective Monoaural Speech Separation through Convolutional Top-Down Multi-View Network

Speech separation, sometimes known as the “cocktail party problem”, is the process of separating individual speech signals from an audio mixture that includes ambient noises and several speakers. The goal is to extract the target speech in this complicated sound scenario and either make it easier to understand or increase its quality so that it may be used in subsequent processing. Speech separation on overlapping audio data is important for many speech-processing tasks, including natural language processing, automatic speech recognition, and intelligent personal assistants. New speech separation algorithms are often built on a deep neural network (DNN) structure, which seeks to learn the complex relationship between the speech mixture and any specific speech source of interest. DNN-based speech separation algorithms outperform conventional statistics-based methods, although they typically need a lot of processing and/or a larger model size. This study presents a new end-to-end speech separation network called ESC-MASD-Net (effective speaker separation through convolutional multi-view attention and SuDoRM-RF network), which has relatively fewer model parameters compared with the state-of-the-art speech separation architectures. The network is partly inspired by the SuDoRM-RF++ network, which uses multiple time-resolution features with downsampling and resampling for effective speech separation. ESC-MASD-Net incorporates the multi-view attention and residual conformer modules into SuDoRM-RF++. Additionally, the U-Convolutional block in ESC-MASD-Net is refined with a conformer layer. Experiments conducted on the WHAM! dataset show that ESC-MASD-Net outperforms SuDoRM-RF++ significantly in the SI-SDRi metric. Furthermore, the use of the conformer layer has also improved the performance of ESC-MASD-Net.

Aircraft Objection Detection Method of Airport Surface based on Improved YOLOv5

An aircraft object detection method on the basis of improved YOLOv5 was proposed to address the issues of large model size, high number of parameters, and inability to meet real-time monitoring requirements of aircrafts in traditional object detection. Firstly, the basic unit of ShuffleNetv2 network was optimized through replacing 3x3 convolution with 5x5 convolution and removing subsequent 1x1 convolution. Simultaneously, the original ReLU activation function was replaced with PReLU. Secondly, CBAM (Convolutional Block Attention Module) attention mechanism was developed to enhance the detection accuracy of the improved network. Finally, improved ShuffleNetv2 network was applied as the backbone structure of YOLOv5. Experimental results revealed that the parameter number of the improved YOLOv5 method introduced in this paper was decreased by 18 times, with a model size of 1.03M. Therefore, a 20.8% increase was achieved in frames per second (FPS) in GPU environments and a 234.6% increase was observed in FPS in CPU environments, while a mean average precision (mAP@0.5) of 0.99 was maintained compared with traditional YOLOv5 network. Because of the advantages of fewer parameters, faster recognition speed, higher localization accuracy, and smaller memory requirement, the developed method was found to be suitable for real-time monitoring of aircrafts in airport surface.

Lightweight Improved YOLOv5s-CGhostnet for Detection of Strawberry Maturity Levels and Counting

A lightweight strawberry detection and localization algorithm plays a crucial role in enabling the harvesting robot to effectively harvest strawberries. The YOLO model has often been used in strawberry fruit detection for its high accuracy, speed, and robustness. However, some challenges exist, such as the requirement for large model sizes, high computation operation, and undesirable detection. Therefore, the lightweight improved YOLOv5s-CGhostnet was proposed to enhance strawberry detection. In this study, YOLOv5s underwent comprehensive model compression with Ghost modules GCBS and GC3, replacing modules CBS and C3 in the backbone and neck. Furthermore, the default GIOU bounding box regressor loss function was replaced by SIOU for improved localization. Similarly, CBAM attention modules were added before SPPF and between the up-sampling and down-sampling feature fusion FPN–PAN network in the neck section. The improved model exhibited higher mAP@0.5 of 91.7% with a significant decrement in model size by 85.09% and a reduction in GFLOPS by 88.5% compared to the baseline model of YOLOv5. The model demonstrated an increment in mean average precision, a decrement in model size, and reduced computation overhead compared to the standard lightweight YOLO models.