An Efficient Multi-Label Classification-Based Municipal Waste Image Identification DOI Open Access
Rongxing Wu, Xingmin Liu, Tian-tian Zhang

и другие.

Processes, Год журнала: 2024, Номер 12(6), С. 1075 - 1075

Опубликована: Май 24, 2024

Sustainable and green waste management has become increasingly crucial due to the rising volume of driven by urbanization population growth. Deep learning models based on image recognition offer potential for advanced classification recycling methods. However, traditional approaches usually rely single-label images, neglecting complexity real-world occurrences. Moreover, there is a scarcity efforts directed at actual municipal data, with most studies confined laboratory settings. Therefore, we introduce an efficient Query2Label (Q2L) framework, powered Vision Transformer (ViT-B/16) as its backbone complemented innovative asymmetric loss function, designed effectively handle multi-label classification. Our experiments newly developed dataset “Garbage In, Garbage Out”, which includes 25,000 street-level each potentially containing up four types waste, showcase Q2L framework’s exceptional ability identify accuracy exceeding 92.36%. Comprehensive ablation experiments, comparing different backbones, functions, substantiate efficacy our approach. model achieves superior performance compared models, mean average precision increase 2.39% when utilizing switching ViT-B/16 improves 4.75% over ResNet-101.

Язык: Английский

The Dual Role of Innovation in Manufacturing: Enhancing Sustainability and Employment Opportunities DOI Creative Commons

Iris Maria Valez Osorio

Organizacija, Год журнала: 2025, Номер 58(1), С. 3 - 19

Опубликована: Фев. 1, 2025

Abstract Background and Purpose The purpose of this study is to investigate how various types innovation impact sustainability measures within manufacturing companies; these include minimizing raw material usage, reducing energy consumption, optimizing waste management. research further evaluates the linkage between job creation, focusing on fosters new employment opportunities enhances in sector. Methodology methodology involves a hierarchical regression analysis conducted sample 1,570 companies Colombia using SPSS software. This approach aims quantitatively assess effectiveness innovation, sustainability, policies industrial organizations. Results findings reveal significant insights into their management environmental sustainability. These results underscore practical implications embracing for long-term benefits, despite immediate costs. Conclusion provides comprehensive examination diverse consequential impacts Additionally, it suggests directions future that could refine enhance practices industry.

Язык: Английский

Процитировано

0

BIOREMEDIATION: A SUPERIOR ALTERNATIVE FOR REMEDIATING TANNERY EFFLUENT-CONTAMINATED SOIL DOI Creative Commons
Aminu Muhammad Gusau, Aminu Yusuf Fardami

FUDMA Journal of Sciences, Год журнала: 2025, Номер 9(2), С. 193 - 208

Опубликована: Фев. 28, 2025

Tannery effluent poses significant risks to soil health, primarily through contamination with heavy metals like chromium, sulphides, and persistent organic pollutants (POPs). These toxic substances inhibit microbial activity, reducing nutrient cycling matter decomposition essential for fertility. Beneficial microorganisms, including nitrogen-fixing bacteria, are particularly affected, leading altered communities dominated by less advantageous, metal-tolerant species. Accumulation of POPs disrupts enzymatic activities, interferes plant root growth, complicates remediation efforts due pollutant migration groundwater potential entry into the food chain. Prolonged exposure such contaminants diminishes fertility, reduces resilience, ecosystem services, posing threats agricultural productivity environmental health. This review was aimed outline what made bioremediation a superior treatment technology among other methods used in remediating tannery contaminated soil. Efforts mitigate impacts involve combination physical, chemical, biological technologies. Physical washing, flushing, thermal desorption focus on removing or isolating contaminants, while chemical approaches as oxidation, reduction, stabilization transform harmful forms immobilize them. Biological leverages microorganisms plants detoxify sustainably. Bioremediation strategies aid bioaugmentation biostimulation do enhance activity address inorganic effectively more than physical methods. Another excellent called phytoremediation can also effectively, Achieving better technique should be coupled stringent industrial regulations, sustainable tanning methods, stakeholder awareness

Язык: Английский

Процитировано

0

Optimizing Waste Management: Reduction Potential Analysis in Pakis Sub-District TPS, Malang Regency DOI Open Access

Rizky Amalia,

Christia Meidiana, Septiana Hariyani

и другие.

IOP Conference Series Earth and Environmental Science, Год журнала: 2025, Номер 1452(1), С. 012034 - 012034

Опубликована: Фев. 1, 2025

Abstract The growing volume of waste presents substantial management issues, especially under the current collect-transport-dispose paradigm, which frequently results in overburdened temporary disposal sites (TDS or TPS) Malang Regency since there is no reduction prior to TDS. Organic waste, as significant fraction was focus study. Mass Balance Analysis used calculate Recovery Factor (RF) TDS for calculating potential four Pakis Sub-district, through scenario. Two scenarios were study focusing on increase, i.e. maintaining RF and changing first second scenario respectively. showed that expanding service area still have same capacity though increasing input. Meanwhile, scenario, increases RF, can decrease 36% transported into These highlight effectiveness maggot-based organic processing achieving targets, offering a scalable model sustainable management.

Язык: Английский

Процитировано

0

A Literature Review on Waste Management Treatment and Control Techniques DOI Creative Commons
Dharmendra Hariyani, Poonam Hariyani, Sanjeev Mishra

и другие.

Sustainable Futures, Год журнала: 2025, Номер unknown, С. 100728 - 100728

Опубликована: Май 1, 2025

Язык: Английский

Процитировано

0

An Efficient Multi-Label Classification-Based Municipal Waste Image Identification DOI Open Access
Rongxing Wu, Xingmin Liu, Tian-tian Zhang

и другие.

Processes, Год журнала: 2024, Номер 12(6), С. 1075 - 1075

Опубликована: Май 24, 2024

Sustainable and green waste management has become increasingly crucial due to the rising volume of driven by urbanization population growth. Deep learning models based on image recognition offer potential for advanced classification recycling methods. However, traditional approaches usually rely single-label images, neglecting complexity real-world occurrences. Moreover, there is a scarcity efforts directed at actual municipal data, with most studies confined laboratory settings. Therefore, we introduce an efficient Query2Label (Q2L) framework, powered Vision Transformer (ViT-B/16) as its backbone complemented innovative asymmetric loss function, designed effectively handle multi-label classification. Our experiments newly developed dataset “Garbage In, Garbage Out”, which includes 25,000 street-level each potentially containing up four types waste, showcase Q2L framework’s exceptional ability identify accuracy exceeding 92.36%. Comprehensive ablation experiments, comparing different backbones, functions, substantiate efficacy our approach. model achieves superior performance compared models, mean average precision increase 2.39% when utilizing switching ViT-B/16 improves 4.75% over ResNet-101.

Язык: Английский

Процитировано

2