Performance Comparison of SRCNN, VDSR, and SRDenseNet Deep Learning Models in Embedded Autonomous Driving Platforms

Kim, Joongheon; Shin, MyungJae; Kim, Dohyun; Park, Soo-Hyun; Kang, Yeongeun; Kim, Jung-Hyun; Lee, Haemin; Yun, Won Joon; Choi, Jae-Ho; Park, Seunghoon; Oh, Seunghyeok; Yoo, Jaesung

doi:10.1109/icoin50884.2021.9333896

Cited by 4 publications

(2 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Dong et al (2015) claimed that the model has the capability to effectively improve the resolution of an image while maintaining the clarity of the image details. The SRCNN is also often used to compare the results of the models (Shi et al, 2016;Kim et al, 2021). the original image).…”

Section: Model Validationmentioning

confidence: 99%

Deep learning-based spatial downscaling and its application for tropical cyclone detection in the western North Pacific

Chen,

Yuan

2024

Front. Earth Sci.

View full text Add to dashboard Cite

Resolution of global climate models (GCMs) significantly influences their capacity to simulate extreme weather such as tropical cyclones (TCs). However, improving the GCM resolution is computationally expensive and time-consuming, making it challenging for many research organizations worldwide. Here, we develop a downscaling model, MSG-SE-GAN, based on the Generative Adversarial Networks (GAN) together with Multiscale Gradient (MSG) technique and a Squeeze-and-Excitation (SE) Net, to achieve 10-folded downscaling. GANs consist of a generator and a discriminator network that are trained adversarially, and are often used for generating new data that resembles a given dataset. MSG enables generation and discrimination of multi-scale images within a single model. Inclusion of an attention layer of SE captures better underlying spatial structure while preserving accuracy. The MSG-SE-GAN is stable and fast converging. It outperforms traditional bilinear interpolation and other deep-learning methods such as Super-Resolution Convolutional Neural Networks (SRCNN) and MSG-GAN in downscaling low-resolution meteorological data in assessment metrics and power spectral density. The MSG-SE-GAN has been used to downscale the TC-related variables in the western North Pacific in the low-resolution GCMs of HadGEM3-GC31 and EC-Earth3P, respectively. The downscaled data show highly similar TC activities to the direct outputs of the high-resolution HadGEM3-GC31 and EC-Earth3P, respectively. These results not only suggest the validity of the MSG-SE-GAN but also indicate its possible portability among low-resolution GCMs.

show abstract

Section: Model Validationmentioning

confidence: 99%

Deep learning-based spatial downscaling and its application for tropical cyclone detection in the western North Pacific

Chen,

Yuan

2024

Front. Earth Sci.

View full text Add to dashboard Cite

show abstract

“…Kim [13] VDSR model, the first residual structure for super-resolution reconstruction. The model network depth reaches 20 layers, and the deeper the network structure has a greater receptive field.…”

Section: Image Super-resolution Based On Deep Learningmentioning

confidence: 99%

Research on an improved fish recognition algorithm based on YOLOX

Liu

Wen²,

Zhou³

2022

ITM Web Conf.

View full text Add to dashboard Cite

The key to the development of underwater resources is to detect underwater targets quickly and accurately in real time. However, due to the influence of light, the underwater image is easy to be distorted and the contrast is low and so on, which greatly affects the performance of the detection algorithm, In order to improve the detection accuracy of underwater targets, After a detailed analysis of the underwater detection target features, The attention mechanism ECA module was added to the YOLOX model, Real-ESRGAN was used to treat multiple target and fuzzy images in detection images, the accuracy improved about 10 percent, A high-precision target detection algorithm suitable for underwater fish was developed, The ideal detection result was achieved.

show abstract