Deep learning image compression, using neural networks, improves compression over traditional methods like JPEG. These methods enhance visual quality at lower bit rates by learning better image representations. However, they struggle with capturing broad context compared to local features. To address this, we propose enhancements: a new convolutional module with stacked layers and advanced operations, and a spatial attention block ("Shuffle attention") for better feature extraction. These boost performance. Our method is faster and requires fewer parameters than state-of-the-art techniques on Kodak and CLIC datasets. Despite slightly lower rate-distortion performance, our Composite Conv module and spatial attention block effectively extract global features and reduce encoding time. In conclusion, our work advances deep learning image compression by mitigating convolutional network limitations, enhancing compression efficiency while preserving quality.
A novel SAR image denoising scheme based on hidden Markov tree (HMT) in the quad-tree complex wavelet packet
transform (QCWPT) domain was presented to achieve the tradeoff between details retainment and noise removal. A
neighborhood coefficient differential window was used to compute intra-scale correlations of complex wavelet
coefficients in high frequency detail subimage, and intra-scale correlational state was identified according to the
smallest error rate Bayesian decision-making rules. A HMT was fitted to describe the correlations between the
complex wavelet coefficients across decomposition scales and mark inter-scale correlational state. The product
results of corresponding positional intra-scale and inter-scale correlational state were looked as a new hidden state
transition probability. A set of iterative equations was developed using the expectation-maximization(EM) algorithm
to estimate the model parameters and produce denoising images. Experimental results show that the proposed
denoising algorithm is superior to the traditional filtering methods and possible to achieve an excellent balance
between suppress speckle noise effectively and preserve as many image details and edges as possible.
The focus of this study is path selection for manufacturing processing, such as finding the shortest processing path, in an application of such a printed circuit board in the electronic industry. This paper models this kind of processing path optimization problem by application of a GA algorithm. First, the related problem of math modeling is discussed, such as coding methods, selection of fitness functions, and choice of genetic operators such as a selection operator, crossover operator, reverse operator, mutation operator and related parameters. All of these are used to build a solving model. Then related factor of genetic optimization algorithm such as initial generation, fitness evaluation, computing steps and so on was designed. The results of simulation and comparisons with practical application show that GA is feasible and valid.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.