Title: |
Predictive Model Based on Machine Learning Techniques for Estimating Blackberry Production in a Region of Mexico |
Authors: |
JAL Méndez, JRG Pulido, JRH Morales |
Source: |
International Journal of Latest Engineering Research and Applications, pp 01 - 10, Vol 10 - No. 12, 2025 |
Abstract: |
Food production, particularly of berries, has experienced a notable increase in recent years, driven by its growing market demand. Mexico has established itself as the world's leading producer of blackberries, with the states of Michoacán and Jalisco standing out for their high production. Within Michoacán, the municipalities of Los Reyes, Peribán, Tacámbaro, Tocumbo, and Ario concentrate the highest production, with Los Reyes being the national leader. |
 |
Download Full Article |
DOI: |
10.56581/IJLERA.10.12.01-10 |
Title: |
A Study on Concrete Filled Steel Tubular Section (CFST) Columns at Composite Structures |
Authors: |
S. Bala Subramanian, G. Karunakaran, G. Manikandan, A. K. Nantha Kumar |
Source: |
International Journal of Latest Engineering Research and Applications, pp 11 - 18, Vol 10 - No. 12, 2025 |
Abstract: |
A RCC and steel frames have been the most common economic frame systems for long times whereas composite frame system has also emerged as popular system for high rise buildings for few decades. Multi-storey composite frames are generally composed of structural steel members made composite with concrete. The use of concrete filled steel tubes (CFST) in building construction has seen renaissance in recent years due to their numerous advantages, apart from its superior structural performance making a typical composite frame structure. Their usage as columns in high-rise and multi-story buildings, as beams in low-rise industrial buildings and as arch bridges, has become extensive in many countries in last four decades with abundant examples. But, their usage in India is a new concept. Hence, this paper shall primarily emphasis to investigate the various aspects of CSFT members in the building industry; primarily considering the various aspects of these members which have turned its unique phase with the advancement of technology. |
Kaywords: |
Keywords: Concrete filled steel tubes, CFST Columns, composite frames and confinement |
|
Download Full Article |
DOI: |
10.56581/IJLERA.10.12.11-18 |
Title: |
A Lightweight Framework with Strip Perception and Adaptive Pooling for Fabric Defect Detection |
Authors: |
Salman Shehzad, Yuan-Gen Wang, Chunsheng Yang, Imran Hussain |
Source: |
International Journal of Latest Engineering Research and Applications, pp 19 - 28, Vol 10 - No. 12, 2025 |
Abstract: |
Fabric surface defects reduce textile quality, so accurate real-time detection is a key to industrial quality control. Current lightweight models such as YOLOv7-Tinier struggle to detect elongated/subtle defects (such as cutting, small holes) while staying efficient for on-site use—limiting factory deployment. To address this, we propose YOLOv7-Tinier-SPAF (Strip Perception and Adaptive Pooling), an improved lightweight framework for textile inspection. In the proposed framework, we add three innovations to YOLOv7-Tinier: (1) Strip Perception Module (SPM), which uses asymmetric kernels to capture long, narrow defects; (2) Squeeze-and-Excitation Spatial Pyramid Pooling-Fast (SESPPF), which combines channel attention and multi-scale pooling to refine features; and (3) Focal-Enhanced Complete Intersection over Union (FECIoU) Loss, which weights hard-to-detect cases to improve bounding box accuracy and defect localization. Evaluated on a textile defect dataset, the experimental results show that YOLOv7-Tinier-SPAF performed better than the Yolov7-Tiny baseline both in performance and speed. Ablation studies confirm that each module works effectively: SPM improves the detection of elongated defects by 18.6%, SE-SPPF improves features by 11.2%, and FECIoU reduces localization errors by 9.4%. It also outperforms models like YOLOv8n both accuracy and robust regression, making it practical for factory textile quality control. |
Kaywords: |
Fabric defect detection, lightweight object detection, YOLOv7-Tinier, strip perception module, adaptive pooling, focal-enhanced CIoU loss. |
|
Download Full Article |
DOI: |
10.56581/IJLERA.10.12.19-28 |
