Search Results (116)

The sizing of photovoltaic (PV) systems has been a concern since the 1990s, particularly with the trend of inverter undersizing as PV module prices decrease. While many studies have assessed the behavior of AC energy and economic parameters with varying Inverter Load Ratios (ILRs), they often neglect the impact of degradation on system lifetime or fail to analyze how it influences ILR selection in depth. This study examines the relationship between DC loss curves and ILRs, their evolution over time, and their effects on efficiency and Final Yield. Simulating solar resources in 27 Brazilian cities, it evaluates clipping losses and optimal ILR values ranging from 0.8 to 2.0 for 28 recent inverters. The research aims to identify the ILR that minimizes the Levelized Cost of Energy (LCOE) while maximizing Final Yield, revealing variations in optimal ILR ranges across different inverter–city combinations. The optimal ILR was between 1.1 and 1.3 for modern medium- and high-power inverters, while low-power inverters had a range of up to 1.8. The findings highlight that practical ILR considerations can overlook real-world challenges, leaving the system’s full potential untapped. Full article

This paper extends our previous work on cleft lip and/or palate (CL/P) classification, which employed vision transformers (ViTs) and Siamese neural networks. We now integrate SigLIP 2, a state-of-the-art multilingual vision–language model, for feature extraction, replacing the previously utilized BiomedCLIP. SigLIP 2 offers enhanced semantic understanding, improved localization capabilities, and multilingual support, potentially leading to more robust feature representations for CL/P classification. We hypothesize that SigLIP 2’s superior feature extraction will improve the classification accuracy of CL/P types (bilateral, unilateral, and palate-only) from the UltraSuite CLEFT dataset, a collection of ultrasound video sequences capturing tongue movements during speech with synchronized audio recordings. A comparative analysis is conducted, evaluating the performance of our original ViT-Siamese network model (using BiomedCLIP) against a new model leveraging SigLIP 2 for feature extraction. Performance is assessed using accuracy, precision, recall, and F1 score, demonstrating the impact of SigLIP 2 on CL/P classification. The new model achieves statistically significant improvements in overall accuracy (86.6% vs. 82.76%) and F1 scores for all cleft types. We discuss the computational efficiency and practical implications of employing SigLIP 2 in a clinical setting, highlighting its potential for earlier and more accurate diagnosis, personalized treatment planning, and broader applicability across diverse populations. The results demonstrate the significant potential of advanced vision–language models, such as SigLIP 2, to enhance AI-powered medical diagnostics. Full article

With the rapid advancement of global digital transformation, enterprises and financial institutions face increasing challenges in managing and processing receipt-like financial documents. Traditional manual document processing methods can no longer meet the demands of modern office operations and business expansion. To address these issues, automated document recognition systems based on computer vision and deep learning technologies have emerged. This paper proposes a receipt recognition technology based on multimodal alignment and lightweight sequence modeling, integrating the CLIP (Contrastive Language-Image Pretraining) and Bidirectional Gated Recurrent Unit (BiGRU) framework. The framework aims to achieve synergistic optimization of image and text information through semantic correction. By leveraging dynamic threshold classification, geometric regression loss, and multimodal feature alignment, the framework significantly improves text detection and recognition accuracy in complex layouts and low-quality images. Experimental results show that the model achieves a detection F1 score of 93.1% and a Character Error Rate (CER) of 5.1% on the CORD dataset. Through a three-stage compression strategy of quantization, pruning, and distillation, the model size is reduced to 18 MB, achieving real-time inference speeds of 25 FPS on the Jetson AGX Orin edge device, with power consumption stabilized below 12 W. This framework provides an efficient, accurate, and edge-computing-friendly solution for automated receipt processing. Practical implications include its potential to enhance the efficiency of financial audits, improve tax compliance, and streamline the operational management of financial institutions, making it a valuable tool for real-world applications in receipt automation. Full article

Hydrogen is expected to play an important role in decarbonizing different heavy industries and the transportation sector. Water electrolysis is, therefore, one of the most rapidly spreading energy technologies, with PEM electrolyzers taking a continuously increasing share in the technology mix. Most often, the aim is to form green hydrogen, utilizing electricity exclusively of renewable origin. The intermittency of such sources, however, poses several technological challenges and financial questions. Focusing on PEM electrolyzers, we discuss the effect of pressure, temperature, and reaction rate changes, induced by the intermittent operation, and general thoughts regarding system component erosion caused by the regular start–stop cycles are also considered. As a case study, we present a high-level techno-economic analysis of data from a pilot 1 MW PEM electrolysis system, coupled to a 20 MW PV farm, deployed in Hungary. We underscore the importance of the often overlooked local regulations and financial incentives, which strongly influence the most beneficial operation scenario. Full article

Large-scale Visual-Language Models have demonstrated powerful adaptability in video recognition tasks. However, existing methods typically rely on fine-tuning or text prompt tuning. In this paper, we propose a visual-only prompting method that employs watermark and trademark prompts to bridge the distribution gap of spatial-temporal video data with Visual-Language Models. Our watermark prompts, designed by a trainable prompt generator, are customized for each video clip. Unlike conventional visual prompts that often exhibit noise signals, watermark prompts are intentionally designed to be imperceptible, ensuring they are not misinterpreted as an adversarial attack. The trademark prompts, bespoke for each video domain, establish the identity of specific video types. Integrating watermark prompts into video frames and prepending trademark prompts to per-frame embeddings significantly boosts the capability of the Visual-Language Model to understand video. Notably, our approach improves the adaptability of the CLIP model to various video action recognition datasets, achieving performance gains of 16.8%, 18.4%, and 13.8% on HMDB-51, UCF-101, and the egocentric dataset EPIC-Kitchen-100, respectively. Additionally, our visual-only prompting method demonstrates competitive performance compared with existing fine-tuning and adaptation methods while requiring fewer learnable parameters. Moreover, through extensive ablation studies, we find the optimal balance between imperceptibility and adaptability. Code will be made available. Full article

In today’s job market, effective social communication is crucial for employment success. We investigated “Cog ‘n’ Role”, a novel video modeling (VM) intervention that integrates video self-modeling (VSM) and social problem-solving therapy (SPST) to enhance socio-vocational skills in individuals with intellectual and developmental disabilities (IDDs). The intervention is delivered via “PowerMod”, an application featuring ready-to-use VM scenarios and enhanced accessibility options; our aim was to examine (a) the app’s social validity and (b) the effectiveness of the intervention in improving job-related social skills. Thirty-four adults with IDD used “PowerMod” to view video clips of common workplace scenarios and rated their experiences through questionnaires. Subsequently, seventeen adults who have social difficulties at work participated in four weekly therapy sessions featuring the “Cog ‘n’ Role” intervention via the PowerMod app. Socio-vocational skills were measured through questionnaires filled out by their counselors; participants found the adapted video clips to be significantly more comprehensible and relevant compared to non-adapted video clips. Additionally, the intervention group showed significant improvements in socio-vocational behaviors and a significant transition to jobs that required higher levels of independence. These findings provide preliminary evidence for the impact of this innovative intervention in enhancing socio-vocational skills among individuals with mild to moderate IDD. Full article

While the free and open-source Broadly Reconfigurable and Expandable Automation Device (BREAD) has demonstrated functionality as an inexpensive replacement for many commercial controllers, some aspects of its design require updating to make it more aligned with commercial supervisory control and data acquisition (SCADA) systems. Some of these updates to BREAD for version 2 included improvements to the mechanical design for stability with an alignment cover, rail mounting with Deutsche Institut für Normung (DIN) rail clips, ESP32 Loaf Controller with local wireless connectivity, and open-source web browser-based software control. These updates were validated by comparing BREAD v2 to an existing commercial controller used for airline-based pH control for industrial seaweed production. BREAD v2 was integrated into an electrical enclosure complete with pH probes, CO₂ lines, solenoid valves, and a power supply. After comparing the two approaches, BREAD v2 was found to be more precise by roughly a factor of five, and less expensive by a factor of three than proprietary systems, while also offering additional functionality like data logging and wireless monitoring. Although able to match or beat specific functions of SCADA systems, future work is needed to transform BREAD into a full SCADA system. Full article

This paper presents an integrated chaos-based algorithm for image encryption that combines the chaotic Hénon map and chaotic logistic map (CLM) to enhance the security of digital image communication. The proposed method leverages chaos theory to generate cryptographic keys, utilizing a 1D key from the logistic map generator and a 2D key from the chaotic Hénon map generator. These chaotic maps produce highly unpredictable and complex keys essential for robust encryption. Extensive experiments demonstrate the algorithm’s resilience against various attacks, including chosen-plaintext, noise, clipping, occlusion, and known-plaintext attacks. Performance evaluation in terms of encryption time, throughput, and image quality metrics validates the effectiveness of the proposed integrated approach. The results indicate that the chaotic Hénon–logistic map integration provides a powerful and secure method for safeguarding digital images during transmission and storage with a key space that reaches up to $2^{200}$. Moreover, the algorithm has potential applications in secure image sharing, cloud storage, and digital forensics, inspiring new possibilities. Full article

The previously proposed cascaded inverse fast Fourier transform/fast Fourier transform (IFFT/FFT)-based point-to-multipoint (P2MP) flexible optical transceivers have the potential to equip future intensity modulation and direct detection (IMDD) optical access networks with excellent flexibility, adaptability, scalability and upgradability. However, due to their cascaded IFFT-based multi-channel aggregations, P2MP flexible transceivers suffer high peak-to-average power ratios (PAPRs). To address the technical challenge, this paper proposes a novel P2MP flexible optical transceiver, which uses a cascaded discrete Fourier transformation-spread (DFT-Spread) IFFT/FFT-based multi-channel aggregation/de-aggregation and standard signal clipping to jointly reduce its PAPRs. The upstream performances of the proposed transceivers are numerically explored in a 20 km IMDD upstream passive optical network (PON). The results indicate that the proposed transceiver’s PAPRs are mainly dominated by the size of the last IFFT operation of the multi-channel aggregation, and are almost independent of modulation format and channel count. Compared to conventional cascaded IFFT/FFT-based P2MP transceivers with and without clipping operations, the proposed DFT-Spread P2MP transceivers can reduce PAPRs by 2.6 dB and 3.5 dB, respectively, for a final IFFT operation size of 1024. More significant PAPR reductions are achievable when the last IFFT operation size is increased further. As a direct result, compared to conventional P2MP transceivers adopting clipping operations only, the proposed transceiver can improve upstream receiver sensitivities by >1.9 dB and the aggregated upstream transmission capacities by >14.1%. Such aggregated upstream transmission capacity enhancements are independent of channel count and become more pronounced for longer transmission distances. Full article

Filtered orthogonal frequency division multiplexing (F-OFDM), employed in visible light communication (VLC) systems, has been considered a promising technique for overcoming OFDM’s large out-of-band emissions and thus reducing bandwidth efficiency. However, due to Hermitian symmetry (HS) imposition, a challenge in VLC involves increasing power consumption and doubling inverse fast Fourier transform IFFT/FFT length. This paper introduces the non-Hermitian symmetry (NHS) Flip-F-OFDM technique to enhance bandwidth efficiency, reduce the peak–average-power ratio (PAPR), and lower system complexity. Compared to the traditional HS-based Flip-F-OFDM method, the proposed method achieves around 50% reduced system complexity and prevents the PAPR from increasing. Therefore, the proposed method offers more resource-saving and power efficiency than traditional Flip-F-OFDM. Then, the proposed scheme is assessed with HS-free Flip-OFDM, asymmetrically clipped optical (ACO)-OFDM, and direct-current bias optical (DCO)-OFDM. Concerning bandwidth efficiency, the proposed method shows better spectral efficiency than HS-free Flip-OFDM, ACO-OFDM, and DCO-OFDM. Full article

The application of DFT precoding in mitigating peak-to-average power ratio (PAPR) issues in optical wireless communication systems under power constraints is well established. However, the channel spatial diversity for multiusers needs to be considered. This paper proposes a joint DFT and spacetime precoding technique to support multiple users and mitigate the PAPR for optical wireless communication systems when OFDM modulation is used. In our study, we introduce the weighted Bussgang theorem as an alternative method for evaluating the communication performance of hard-clipped systems due to the transmitted power constraint. Our numerical results demonstrate the effectiveness of this approach in analyzing SNR and BER performance. Full article

Crowd behavior presents significant challenges due to intricate interactions. This research proposes an approach that combines the power of 3D Convolutional Neural Networks (ConvNet) and Multi-Support Vector Machines (Multi-SVM) to study and analyze crowd behavior in highly dense crowd videos. The proposed approach effectively utilizes the temporal information captured by the 3D ConvNet, which accounts for the spatiotemporal characteristics of crowd movement. By incorporating the third dimension as a temporal stack of images forming a clip, the network can learn and comprehend the dynamics and patterns of crowd behavior over time. In addition, the learned features from the 3D ConvNet are classified and interpreted using Multi-SVM, enabling a comprehensive analysis of crowd behavior. This methodology facilitates the identification and categorization of various crowd dynamics, including merging, diverging, and dense flows. To evaluate the effectiveness of the approach, experiments are conducted on the Crowd-11 dataset, which comprises over 6000 video sequences with an average length of 100 frames per sequence. The dataset defines a total of 11 crowd motion patterns. The experimental results demonstrate promising recognition rates and achieve an accuracy of 89.8%. These findings provide valuable insights into the complex dynamics of crowd behavior, with potential applications in crowd management. Full article

Drowsy driving is a leading cause of commercial vehicle traffic crashes. The trend is to train fatigue detection models using deep neural networks on driver video data, but challenges remain in coarse and incomplete high-level feature extraction and network architecture optimization. This paper pioneers the use of the CLIP (Contrastive Language-Image Pre-training) model for fatigue detection. And by harnessing the power of a Transformer architecture, sophisticated and long-term temporal features are adeptly extracted from video sequences, paving the way for more nuanced and accurate fatigue analysis. The proposed CT-Net (CLIP-Transformer Network) achieves an AUC (Area Under the Curve) of 0.892, a 36% accuracy improvement over the prevalent CNN-LSTM (Convolutional Neural Network-Long Short-Term Memory) end-to-end model, reaching state-of-the-art performance. Experiments show that the CLIP pre-trained model more accurately extracts facial and behavioral features from driver video frames, improving the model’s AUC by 7% over the ImageNet-based pre-trained model. Moreover, compared with LSTM, the Transformer more flexibly captures long-term dependencies among temporal features, further enhancing the model’s AUC by 4%. Full article

In light detection and ranging (LiDAR) applications, correct intensities from echo data intuitively contribute to the characterization of target reflectivity. However, the power in raw echo waveforms may be clipped owing to the limited dynamic range of LiDAR sensors, which directly results in false intensity values generated by existing LiDAR systems working in scenarios involving highly reflective objects or short distances. To tackle the problem, an asymmetric Gaussian echo model is proposed in this paper so as to recover echo power–time curves faithfully to its optical physics. Considering the imbalance in temporal length and steepness between rising and falling edges, the echo model features a shared mean and two distinct standard deviations on both sides. The accuracy and effectiveness of the proposed model are demonstrated by correcting the power–time curve from a real LiDAR loaded with avalanche photodiode (APD) sensors and estimating the reflectivities of real targets. As when tested by targets with reflectivities from low to high placed at distances from near to far, the model achieves a maximum of 41.8-fold improvement in relative error for the same target with known reflectivity and a maximum of 36.0-fold improvement in the coefficient of variation for the same target along the whole range of 100 m. Providing accurate and stable characterization of reflectivity in different ranges, the model greatly boosts applications consisting of semantic segmentation and object recognition, such as autonomous driving and environmental monitoring. Full article

Maximizing energy efficiency (EE) in massive multiple-input multiple-output (MIMO) systems, while supporting the rapid expansion of Internet of Things (IoT) devices, is a critical challenge. In this paper, we delve into the intricate operations geared toward enhancing EE in such complex environments. To effectively support a multitude of IoT devices, we adopt a strategy of heavy reference signal (RS) reuse, and in this circumstance, we formulate the EE metrics and their corresponding inverses to determine pivotal operational parameters. These EE-centric parameters encompass factors such as the number of service antennas in the base station (BS), the number of IoT devices, and permissible coverage extents. Our objective is to calibrate these parameters to meet a predefined EE threshold, ensuring optimal system performance. Additionally, we recognize the indispensable role of Peak-to-Average Power Ratio (PAPR) reduction techniques, particularly in multicarrier systems, to further enhance EE. As such, we employ clipping-based PAPR reduction methods to mitigate signal distortions and bolster overall efficiency. Theoretical EE metrics are derived based on formulated signal-to-interference-plus-noise ratios (SINRs), yielding insightful closed-form expressions for the operational parameters. Leveraging two distinct EE metric models, we undertake parameter determinations, accounting for the levels of approximation. Intriguingly, our analysis reveals that even simplified models exhibit remarkable applicability in real-world scenarios, with a minimal margin of error. The results not only underscore the practical applicability of our theoretical constructs but also highlight the potential for significant EE enhancements in massive MIMO systems, thereby contributing to sustainable evolution in the IoT era. Full article