► Classify all of the pixels in an image into several clusters according to their colors. ► By measuring the spatial distance among the pixels in a same cluster, the three types of color spatial ...distribution (CSD) features of the image is obtained. ► Based on the three types of CSD features, three image retrieval methods are also provided. ► To accelerate the image retrieval methods, a fast filter is also presented to eliminate most undesired images in advance. ► A genetic algorithm is also given to decide the most suitable parameters which are used in the proposed image retrieval methods.
In this paper, three types of image features are proposed to describe the color and spatial distributions of an image. In these features, the
K-means algorithm is adopted to classify all of the pixels in an image into several clusters according to their colors. By measuring the spatial distance among the pixels in a same cluster, the three types of color spatial distribution (CSD) features of the image is obtained. Based on the three types of CSD features, three image retrieval methods are also provided. To accelerate the image retrieval methods, a fast filter is also presented to eliminate most undesired images in advance. A genetic algorithm is also given to decide the most suitable parameters which are used in the proposed image retrieval methods. The proposed image retrieval methods are simple. Moreover, the experiments show that the proposed methods can provide impressive results as well.
► In this study, an NCC detector is proposed to automatically detect the cytoplast and nucleus contours of cell in a cervical smear image. ► The NCC detector employs the adaptable threshold decision ...(ATD) method to separate the cell from the cervical smear image, and then uses the maximal gray-level-gradient-difference (MGLGD) method to extract the nucleus from the cell. ► Experimental results tell that the NCC detector is better than GVF-ACM and ENNCC detector in segmenting the cytoplast and nucleus of a cell.
In this paper, a nucleus and cytoplast contour detector (NCC detector) is presented to automatically detect the cytoplast and nucleus contours of a cell in a cervical smear image. The NCC detector uses the adaptable threshold decision (ATD) method to separate the cell from the cervical smear image, and then uses the maximal gray-level-gradient-difference (MGLGD) method, proposed in this paper, to extract the nucleus from the cell. The experimental results show that the NCC detector is superior to two existing methods, the gradient vector flow-active contour model (GVF-ACM) and the edge enhancement nucleus and cytoplast contour (ENNCC) detector, in segmenting the cytoplast and nucleus of a cell.
In traditional VSS schemes, the size of the share image is substantially expanded since each pixel of the secret image is mapped onto a block consisting of several pixels. In addition, the quality of ...the reconstructed secret image is normally degraded in contrast, especially for halftone images. This study proposes a VSS scheme that maps a block in a secret image onto one corresponding equal-sized block in each share image without image size expansion. Two types of techniques, including histogram width-equalization and histogram depth-equalization, are proposed to generate the corresponding share blocks containing multiple levels rather than two levels based on the density of black pixels on the blocks for a secret block. In the former technique, the gray-scale image histogram is obtained by uniformly splitting the range of the pixel gray levels in the secret image, while in the latter the buckets are created so that the area of each bucket is roughly constant by containing approximately the same number of pixels. The proposed schemes significantly improve the quality of the reconstructed secret image compared to several previous investigations.
This paper is intended to present a lossless image compression method based on multiple-tables arithmetic coding (MTAC) method to encode a gray-level image f. First, the MTAC method employs a median ...edge detector (MED) to reduce the entropy rate of f. The gray levels of two adjacent pixels in an image are usually similar. A base-switching transformation approach is then used to reduce the spatial redundancy of the image. The gray levels of some pixels in an image are more common than those of others. Finally, the arithmetic encoding method is applied to reduce the coding redundancy of the image. To promote high performance of the arithmetic encoding method, the MTAC method first classifies the data and then encodes each cluster of data using a distinct code table. The experimental results show that, in most cases, the MTAC method provides a higher efficiency in use of storage space than the lossless JPEG2000 does.
•Efficient CNN architecture search for object classification.•Cost-Effective hardware implementation of lightweight CNN architectures.•Novel controller model.•Enhanced efficiency and practicality.
...Determining the architecture of deep learning models is a complex task. Several automated search techniques have been proposed, but these methods typically require high-performance graphics processing units (GPUs), manual parameter adjustments, and specific training approaches. This study introduces an efficient, lightweight convolutional neural network architecture search approach tailored for object classification. It features an optimized search space design and a novel controller design.
This study introduces a refined search space design incorporating optimizations in both spatial and operational aspects. The focus is on the synergistic integration of convolutional units, dimension reduction units, and the stacking of Convolutional Neural Network (CNN) architectures. To enhance the search space, ShuffleNet modules are integrated, reducing the number of parameters and training time. Additionally, BlurPool is implemented in the dimension reduction unit operation to achieve translational invariance, alleviate the gradient vanishing problem, and optimize unit compositions. Moreover, an innovative controller model, Stage LSTM, is proposed based on Long Short-Term Memory (LSTM) to generate lightweight architectural sequences. In conclusion, the refined search space design and the Stage LSTM controller model are synergistically combined to establish an efficient and lightweight architecture search technique termed Stage and Lightweight Network Architecture Search (SLNAS).
The experimental results highlight the superior performance of the optimized search space design, primarily when implemented with the Stage LSTM controller model. This approach shows significantly improved accuracy and stability compared to random, traditional LSTM, and Genetic Algorithm (GA) controller models, with statistically significant differences. Notably, the Stage LSTM controller excels in accuracy while producing models with fewer parameters within the expanded architecture search space.
The study adopts the Stage LSTM controller model due to its ability to approximate optimal sequence structures, particularly when combined with the optimized search space design, referred to as SLNAS. SLNAS's performance is evaluated through experiments and comparisons with other Neural Architecture Search (NAS) and object classification methods from different researchers. These experiments consider model parameters, hardware resources, model stability, and multiple datasets. The results show that SLNAS achieves a low error rate of 2.86 % on the CIFAR-10 dataset after just 0.2 days of architecture search, matching the performance of manually designed models but using only 2 % of the parameters. SLNAS consistently demonstrates robust performance across various image classification domains, with an approximate parameter count 700,000.
To summarize, SLNAS emerges as a highly effective automated network architecture search method tailored for image classification. It streamlines the model design process, making it accessible to researchers without specialized knowledge in deep learning. Optimizing this method unlocks the full potential of deep learning across diverse research areas. Interested parties can publicly access the source code and pre-trained models through the following link: https://github.com/huanyu-chen/LNASG-and-SLNAS-model.
An adaptable threshold detector Pai, Pei-Yan; Chang, Chin-Chen; Chan, Yung-Kuan ...
Information sciences,
04/2011, Letnik:
181, Številka:
8
Journal Article
Recenzirano
A data set often comprises some data classes. For example, a gray-scale image may consist of some objects, each of which has similar pixels’ gray-scales. The threshold obtained by Otsu’s thresholding ...method (OTM) is biased towards certain data class with larger variance or larger number of data when the variances or the numbers of data among classes are quite different. In this paper, Adaptable Threshold Detector (ATD) is proposed to improve the effectiveness of OTM in determining proper thresholds by dividing class variance by class interval. ATD is more versatile at selecting application-dependent thresholds by changing two parameter values which describe the relative importance among data size, standard deviation, and class interval of a class. In this paper, ATD is applied to crop the expected objects from images to verify its effect upon thresholding. Experimental results demonstrate that ATD is able to perform better than OTM in segmenting objects from images, besides excelling over the Valley-Emphasis Method (VEM) and the Minimum Class Variance Thresholding Method (MCVTM). ATD is also suitable for separating objects from serialized video images, i.e. computerized tomography.
Abstract
Background:
Immunotherapy, especially anti-programmed cell death protein 1 (PD-1) antibodies, has yielded significant and durable tumor response in melanoma, renal cell carcinoma, and other ...cancer types. In contrast, immunotherapy applied major breakthrough achievement only on patients with microsatellite instability high (MSI-H) metastatic colorectal cancer (mCRC). However, MSI-H mCRC comprised only 1.8-4% of total mCRC patients. It is crucial to investigate immune pathways to develop a new strategy for immunotherapy in treatment of microsatellite stable (MSS) mCRC. Recently, some evidences indicate that interferon (IFN)-γ pathway is critical for anti-PD-1 therapy. In this study, we emphasized on evaluating the response of MSS CRC cell lines to IFN-γ.
Methods:
We characterized two MSS CRC cell lines, namely SW480 (KRAS G12V mutation, BRAF wild type) and COLO320 (KRAS wild type, BRAF wild type) for our studies. The impacts of interferon-γ (IFN-γ) on cell surface expressions of different isotypes of major histocompatibility complex (MHC) class I and MHC class 1 related molecule A/B (MICA/B, the NK cell ligand) were explored. Different concentration of IFN-γ with different incubation time were also examined.
Results:
Both cell lines demonstrated low baseline human leukocyte antigen (HLA)-ABC and HLA-BC expression and the expression significantly increased in response to IFN-γ stimulation. We further focused on the SW480 cell line. The SW480 demonstrated low expression of all MHC class I isotypes, including HLA-ABC, HLA-BC, HLA-A, HLA-C, HLA-E, HLA-F, HLA-G and NK cell ligand MICA/B. The IFN-γ specifically stimulated the expression of HLA-A, but all other isotypes were not responsive to IFN-γ stimulation. In summary, IFN-γ significantly stimulates the expression of HLA-ABC, HLA-BC and HLA-A without stimulating HLA-C. The NK cell ligand MIC A/B was also not responsive to IFN-γ stimulation. The stimulatory effect of HLA-ABC and HLA-A in response to IFN-γ stimulation was positively correlated with IFN-γ dosage, which demonstrated highest expression level in response to 800U/ml IFN-γ stimulation. The stimulatory effect in response to IFN-γ stimulation was also positively correlated with increased incubation time with IFN-γ.
Conclusion:
These findings reveal that regardless of their KRAS mutation status, both MSS CRC cell lines possess a phenotype that may be not responsive to anti-PD-1 therapy due to
low expression of MHC class I molecules. The IFN-γ specifically elicits the expression of stimulatory isotype, HLA-A and HLA-B, without eliciting all other inhibitory MHC class I isotypes. These results indicated that IFN-γ demonstrated pure adaptive immune stimulatory effect on CRC cell lines. Analysis of T- and NK cell-responsive immune markers along with IFN-γ signaling pathway may help us to survey possible combination therapy with anti-PD-1 treatment in CRC.
Citation Format: Yi-Hsin Liang, Kuan-Yu Chan, Chang-Cheng Lee, Te-Jung Chen, Ann-Lii Cheng, Kun-Huei Yeh. IFN-γ elicits stimulatory MHC class I isotypes in human colorectal carcinoma cell lines with genetic features of microsatellite stable abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 608.
A novel image feature called color variances among adjacent objects (CVAAO) is proposed in this study. Characterizing the color variances between contiguous objects in an image, CVAAO can effectively ...describe the principal colors and texture distribution of the image and is insensitive to distortion and scale variations of images. Based on CVAAO, a CVAAO-based image retrieval method is constructed. When given a full image, the CVAAO-based image retrieval method delivers the database images most similar to the full image to the user. This paper also presents a CVAAO-based ROI image retrieval method. When given a clip, the CVAAO-based ROI image retrieval method submits to the user a database image containing a target region most similar to the clip. The experimental results show that the CVAAO-based ROI image retrieval method can offer impressive results in finding out the database images that meet user requirements.