• ETRI Journal
• /
• 제43권6호
• /
• pp.991-1003
• /
• 2021

Automatic modulation classification is essential in radar emitter identification. We propose a cascade classifier by combining a support vector machine (SVM) and convolutional neural network (CNN), considering that noise might be taken as radar signals. First, the SVM distinguishes noise signals by the main ridge slice feature of signals. Second, the complex envelope features of the predicted radar signals are extracted and placed into a designed CNN, where a modulation classification task is performed. Simulation results show that the SVM-CNN can effectively distinguish radar signals from noise. The overall probability of successful recognition (PSR) of modulation is 98.52% at 20 dB and 82.27% at -2 dB with low computation costs. Furthermore, we found that the accuracy of intermediate frequency estimation significantly affects the PSR. This study shows the possibility of training a classifier using complex envelope features. What the proposed CNN has learned can be interpreted as an equivalent matched filter consisting of a series of small filters that can provide different responses determined by envelope features.

• Journal of Information Processing Systems
• /
• 제17권2호
• /
• pp.385-398
• /
• 2021

The complexity of deep learning models affects the real-time performance of gesture recognition, thereby limiting the application of gesture recognition algorithms in actual scenarios. Hence, a residual learning neural network based on a deep convolutional neural network is proposed. First, small convolution kernels are used to extract the local details of gesture images. Subsequently, a shallow residual structure is built to share weights, thereby avoiding gradient disappearance or gradient explosion as the network layer deepens; consequently, the difficulty of model optimisation is simplified. Additional convolutional neural networks are used to accelerate the refinement of deep abstract features based on the spatial importance of the gesture feature distribution. Finally, a fully connected cascade softmax classifier is used to complete the gesture recognition. Compared with the dense connection multiplexing feature information network, the proposed algorithm is optimised in feature multiplexing to avoid performance fluctuations caused by feature redundancy. Experimental results from the ISOGD gesture dataset and Gesture dataset prove that the proposed algorithm affords a fast convergence speed and high accuracy.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권12호
• /
• pp.6018-6033
• /
• 2018

Fires are a common cause of catastrophic personal injuries and devastating property damage. Every year, many fires occur and threaten human lives and property around the world. Providing early important sign for early fire detection, and therefore the detection of smoke is always the first step in fire-alarm systems. In this paper we propose an automatic smoke detection system built on camera surveillance and image processing technologies. The key features used in our algorithm are to detect and track smoke as moving objects and distinguish smoke from non-smoke objects using a convolutional neural network (CNN) model for cascade classification. The results of our experiment, in comparison with those of some earlier studies, show that the proposed algorithm is very effective not only in detecting smoke, but also in reducing false positives.

• Current Optics and Photonics
• /
• 제6권1호
• /
• pp.92-103
• /
• 2022

The development of midinfrared (mid-IR) quantum cascade lasers (QCLs) has enabled rapid high-contrast measurement of the mid-IR spectra of biological tissues. Several studies have compared the differences between the mid-IR spectra of colon cancer and noncancerous colon tissues. Most mid-IR spectrum classification studies have been proposed as machine-learning-based algorithms, but this results in deviations depending on the initial data and threshold values. We aim to develop a process for classifying colon cancer and noncancerous colon tissues through a deep-learning-based convolutional-neural-network (CNN) model. First, we image the midinfrared spectrum for the CNN model, an image-based deep-learning (DL) algorithm. Then, it is trained with the CNN algorithm and the classification ratio is evaluated using the test data. When the tissue microarray (TMA) and routine pathological slide are tested, the ML-based support-vector-machine (SVM) model produces biased results, whereas we confirm that the CNN model classifies colon cancer and noncancerous colon tissues. These results demonstrate that the CNN model using midinfrared-spectrum images is effective at classifying colon cancer tissue and noncancerous colon tissue, and not only submillimeter-sized TMA but also routine colon cancer tissue samples a few tens of millimeters in size.

• 전기전자학회논문지
• /
• 제21권4호
• /
• pp.388-396
• /
• 2017

얼굴 검출에는 다양한 포즈, 빛의 세기, 얼굴이 가려지는 현상 등의 많은 변수가 존재하므로, 높은 성능의 검출 시스템이 요구된다. 이에 영상 분류에 뛰어난 Convolutional Neural Network (CNN)이 적절하나, CNN의 많은 연산은 고성능 하드웨어 자원을 필요로한다. 그러나 얼굴 검출을 위한 소형, 모바일 시스템의 개발에는 저가의 저전력 환경이 필수적이고, 이를 위해 본 논문에서는 소형의 FPGA를 타겟으로, 얼굴 검출에 적절한 3-Stage Cascade CNN 구조를 기반으로하는 CPU-FPGA 통합 시스템을 설계 구현한다. 가속을 위해 알고리즘 단계에서 Adaptive Region of Interest (ROI)를 적용했으며, Adaptive ROI는 이전 프레임에 검출된 얼굴 영역 정보를 활용하여 CNN이 동작해야 할 횟수를 줄인다. CNN 연산 자체를 가속하기 위해서는 FPGA Accelerator를 이용한다. 가속기는 Bottleneck에 해당하는 Convolution 연산의 가속을 위해 FPGA 상에 다수의 FeatureMap을 한번에 읽어오고, Multiply-Accumulate (MAC) 연산을 병렬로 수행한다. 본 시스템은 Terasic사의 DE1-SoC 보드에서 ARM Cortex A-9와 Cyclone V FPGA를 이용하여 구현되었으며, HD ($1280{\times}720$)급 입력영상에 대해 30FPS로 실시간 동작하였다. CPU-FPGA 통합 시스템은 CPU만을 이용한 시스템 대비 8.5배의 전력 효율성을 보였다.