The non-local module works as a particularly useful technique for semantic segmentation while criticized for its prohibitive computation and GPU memory occupation. In this paper, we present ...Asymmetric Non-local Neural Network to semantic segmentation, which has two prominent components: Asymmetric Pyramid Non-local Block (APNB) and Asymmetric Fusion Non-local Block (AFNB). APNB leverages a pyramid sampling module into the non-local block to largely reduce the computation and memory consumption without sacrificing the performance. AFNB is adapted from APNB to fuse the features of different levels under a sufficient consideration of long range dependencies and thus considerably improves the performance. Extensive experiments on semantic segmentation benchmarks demonstrate the effectiveness and efficiency of our work. In particular, we report the state-of-the-art performance of 81.3 mIoU on the Cityscapes test set. For a 256x128 input, APNB is around 6 times faster than a non-local block on GPU while 28 times smaller in GPU running memory occupation. Code is available at: https://github.com/MendelXu/ANN.git.
Display omitted
•Two different structures of double layers SSPCMs are presented.•The new wall system is active whole year couple with air-conditioner.•Melting temperatures of two types of SSPCM walls ...are optimized.•Energy saving potential of two types of SSPCM rooms are given.
Double layers SSPCM wall could improve indoor thermal performance and reduce building energy consumption for a whole year. SSPCM layer with higher melting temperature was active in hot seasons and that with lower melting temperature was active in cold seasons. One SSPCM layer on external and the other one on internal wall surface was named “SSPCM wall I”. Two SSPCM layers on internal surface was named “SSPCM wall II”. Location of SSPCM layer affected heat transfer process of building envelope. In order to improve performance of double layers SSPCM wall by optimal location, two dynamic heat transfer models of double layers SSPCM walls in Wuhan city were developed by TRNSYS. The results showed that room with SSPCM wall I could reduce energy consumption 20.51% in winter and 0.65% in summer compared with room with reference wall I. And room with SSPCM wall II could reduce energy consumption 18.75% in winter and 13.46% in summer compared with room with reference wall II. The indoor air temperatures of rooms with SSPCM wall I and II were all in comfortable range for a random selected day both in winter and summer. The optimum melting temperature of SSPCM layer was close to set point of indoor air temperature. It was 0–1°C higher than set point of indoor air temperature in winter and 0–2°C lower than set point of indoor air temperature in summer.
PDF
