Search

With the continuous development of deep learning and image segmentation, image panoptic segmentation has become a research hotspot in the field of computer vision, and many image panoptic segmentation methods have been proposed. This paper summarizes the research methods of image panoptic segmentation based on deep learning. Firstly, the research status of image panoptic segmentation at home and abroad is introduced, and the existing image panoptic segmentation methods are classified according to different optimization tasks in the network architecture, mainly including image panoptic segmentation optimized by feature extraction, image panoptic segmentation optimized by sub-task segmentation, image panoptic segmentation optimized by sub-task fusion, and other image panoptic segmentation. Secondly, 5 commonly used datasets, i.e. MS COCO, PASCAL VOC, Cityscapes, ADE20K and Mapillary Vistas, and 2 evaluation criteria, i.e. panoptic quality (PQ) and parsing covering (PC) in image panoptic segmentation are briefly introduced. And then, performance comparison of typical image panoptic segmentation methods has been conducted on different datasets. Thirdly, the application of image panoptic segmentation in medicine, autonomous driving, drones, agriculture, animal husbandry, military and other fields are listed. Finally, the deficiencies and challenges of existing methods in complex scene applications, real-time performance, and conflicts are pointed out, and the potential directions of image panoptic segmentation are discussed, including image panoptic segmentation based on a simple unified framework, real-time high-quality image panoptic segmentation, and image panoptic segmentation in complex application scenarios.

Climate change can significantly affect building energy use and associated greenhouse gas (GHG) emissions in urban areas, as fossil fuels remain a significant energy source. Green roofs can offer multiple benefits to the urban environment, but their effects on GHG mitigation have not been fully investigated, especially under climate change. This study assessed green roofs' contribution to GHG mitigation by saving building energy and absorbing CO2 under the present (2017–2019) and future (2049–2051) climate scenarios (SSP2-45 and SSP5-85) in Shanghai, China, at the city and township scale. A Geographic Information System (GIS)-based spatial statistical method was developed based on climate change modeling and building energy simulation. The results suggested that installing green roofs can effectively save building energy regardless of building type, yet the amount of savings can vary depending on the weather conditions within the city. The contribution analysis indicated that most saved building energy was attributed to the Heating, Ventilation, and Cooling (HVAC) system, with more energy saved under warmer climate scenarios in the future, particularly during the summer months. More energy was saved from shopping malls on an annual and monthly scale, regardless of the climate scenarios and weather zones. Finally, a case study indicated installing green roofs on all five types of buildings (office, hotel, hospital, shopping mall, apartment) of less than 50 m in height can reduce 8.28% of the CO2 emitted during the building operation stage in the entire city under the present climate scenario. The annual CO2 reduction varied with the location of townships, ranging from 2.18% to 13.78%, depending on the composition of building types and local weather conditions in Shanghai. This study offered policymakers a reference on the environmental benefits and investment values of installing green roofs in large cities. [ABSTRACT FROM AUTHOR]

The energy demand and associated greenhouse gas (GHG) emissions of buildings are significantly affected by the characteristics of the building and local climate conditions. While energy use datasets with high spatial and temporal resolution are highly needed in the context of climate change, energy use monitoring data are not available for most cities. This study introduces an approach combining building energy simulation, climate change modeling, and GIS spatial analysis techniques to develop an energy demand data inventory enabling assessment of the impacts of climate change on building energy consumption in Shanghai, China. Our results suggest that all types of buildings exhibit a net increase in their annual energy demand under the projected future (2050) climate conditions, with the highest increase in energy demand attributed to Heating, Ventilation, and Cooling (HVAC) systems. Variations in building energy demand are found across building types. Due to the large number of residential buildings, they are the main contributor to the increases in energy demand and associated CO2 emissions. The hourly residential building energy demand on a typical hot summer day (29 July) under the 2050 climate condition at 1 p.m. is found to increase by more than 40%, indicating a risk of energy supply shortage if no actions are taken. The spatial pattern of total annual building energy demand at the individual building level exhibited high spatial heterogeneity with some hotspots. This study provides an alternative method to develop a building energy demand inventory with high temporal resolution at the individual building scale for cities lacking energy use monitoring data, supporting the assessment of building energy and GHG emissions under both current and future climate scenarios at minimal cost. [ABSTRACT FROM AUTHOR]

Domain Adaptive Object Detection (DAOD) alleviates the challenge of labor-intensive annotations by transferring semantic information from a labeled source domain to an unlabeled target domain. However, the DAOD suffers from biased discrimination and negative transfer in the thermal domain due to the inherent heterogeneity between the RGB and thermal images. To address the above issues, we propose the Unbiased Granularity Alignment (UGA) framework to facilitate the unified alignment for RGB-Thermal DAOD. Specifically, we devise a Channel Self-encoding Adaptation (CSA) module to mitigate biased discrimination from the discriminative enhancement perspective. CSA aligns the intra-domain channel subspace for inter-domain channel harmonizing. Upon revisiting instance alignment, we uncovered inaccuracies proposals and unstable positive sample phenomena. Therefore, we propose the Relative Relationship Adaptation (RRA) module to mitigate negative transfer. RRA ensures inter-domain semantic consistency through sparse instance alignment. Extensive experiments are conducted on visible-to-thermal and visible-to-visible benchmarks to validate the effectiveness, and our UGA framework outperforms state-of-the-art by a remarkable margin. The code of our UGA is available at https://github.com/zyfone/UGA. [ABSTRACT FROM AUTHOR]

Building energy consumption is an essential source of greenhouse gas (GHG) and air pollution. Green roofs can directly absorb ambient CO2 and remove air pollutants through their vegetation layers, but a limited number of studies have examined their effects on GHG and air pollutant reduction associated with building energy savings, especially in the context of climate change. This research examined the performance of green roofs on CO2 and air pollutant reduction, including SO2, PM2.5, and NOx, through building energy demand savings in Shanghai, China. Climate change mitigation effects were assessed based on the energy consumption of five types of buildings before and after the installation of green roofs under 2020 and 2050 climate conditions, respectively. EnergyPlus software 9.5.0 was applied to simulate hourly energy consumption for different building prototypes with and without green roofs. Green roofs on all building types exhibited positive energy savings on annual, monthly, and diurnal scales, and they can save more energy for most of the building types under the projected 2050 climate condition. Moreover, most of the building energy saved by green roofs came from the Heating, Ventilation, and Cooling (HVAC) systems. In addition, this study discovered that the energy-saving benefits of green roofs vary based on the type of building they were installed on. Green roofs were found to have the largest energy saving on the shopping mall, especially on extremely hot summer days. Finally, a Geographic Information System (GIS)-based approach was developed with the ability to quantify the amount of GHG and air pollutant reduction associated with building energy savings for existing buildings in the Huangpu District of Shanghai. This approach was also utilized to present the spatial distribution of buildings with different levels of suitability to install green roofs by considering their location attributes and air pollutant reduction potential together, which is the major innovation of this research. The purpose of this study is to provide valuable guidance to policy makers regarding the performance of green roofs in building energy-saving and air quality improvement in the urban environment when facing the challenge of climate change, which is essential for urban sustainability. [ABSTRACT FROM AUTHOR]