• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2019.06a
• /
• pp.84-85
• /
• 2019

MPEG-NNR (Compression of Neural Network for Multimedia Content Description and Analysis) aims to define a compressed and interoperable representation of trained neural networks. In this paper, we present a low-rank approximation to compress a CNN used for image compression, which is one of MPEG-NNR use cases. In the presented method, the low-rank approximation decomposes one 2D kernel matrix of weights into two 1D kernel matrix values in each convolution layer to reduce the data amount of weights. The evaluation results show that the model size of the original CNN is reduced to half as well as the inference runtime is reduced up to about 30% with negligible loss in PSNR.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2020.07a
• /
• pp.662-663
• /
• 2020

최근 MPEG 에서는 인공신경망 모델을 다양한 딥러닝 프레임워크에서 상호운용 가능한 포맷으로 압축 표현할 수 있는 NNR(Compression of Neural Network for Multimedia Content Description and Analysis) 표준화를 진행하고 있다. 본 논문에서는 MPEG-NNR 에서 CNN 모델을 압축하기 위한 지역 비선형 양자화(Local Non-linear Quantization: LNQ) 기법을 제시한다. 제안하는 LNQ 는 균일 양자화된 CNN 모델의 각 계층의 가중치 행렬 블록 단위로 추가적인 비선형 양자화를 적용한다. 또한, 제안된 LNQ 는 가지치기(pruning)된 모델의 경우 블록내의 영(zero) 값의 가중치들은 그대로 전송하고 영이 아닌 가중치만을 이진 군집화를 적용한다. 제안 기법은 음성 분류를 위한 CNN 모델(DCASE Task)의 압축 실험에서 기존 균일 양자화를 대비 동일한 분류 성능에서 약 1.78 배 압축 성능 향상이 있음을 확인하였다.

• Korean Journal of Plant Resources
• /
• v.21 no.1
• /
• pp.103-109
• /
• 2008

Nelumbo nucifera root(NNR) is used to clear summerheat(暑熱), bear Yang(陽) upwards and stop bleeding as mentioned in traditional Korean medicine. Also, it has been known that NNR is effective for lowering blood pressure and hyperlipidemia. The rhizome is considered to be nutritive, demulcent, diuretic and cholagogue and is used to treat piles, dyspepsia and diarrhea. An increasingly growing market for nutraceuticals and functional foods has triggered the study on natural sources for nutraceuticals, health foods and functional foods. But rhizome was inconvenient to formulate liquid dosage form(extract) by way of hot water because of its limited storage. Also the majority of the consumers have a complaint against the dosage. The purpose of this study was to develop the functional materials from NNR without side effects. We formulated the solid dosage form viz tablet and granule from the lotus root. Sensory evaluation was performed in terms of smell, taste, color and overall of lotus root and all colored forms(brown, dark brown, light green and yellow) of tablet and granule to evaluate the acceptability of the formulated tablets and granules. In sensory evaluation, among the formulated tablets and granules, light green granules obtained best score overall and yellow tablets showed the overall improved acceptability. In conclusion, lotus rhizome could be recommended as functional food. Further studies to clarify bioactive functions of Nelumbo nucifera in experimental animal model on atopic dermatitis are in progress.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2018.11a
• /
• pp.133-134
• /
• 2018

MPEG-NNR (Compressed Representation of Neural Networks) aims to define a compressed and interoperable representation of trained neural networks. In this paper, a compressed representation of NN and its evaluation performance along with use cases of image/video compression in MPEG-NNR are presented. In the compression of NN, a CNN to replace the in-loop filter in VVC (Versatile Video Coding) intra coding is compressed by applying uniform quantization to reduce the trained weights, and the compressed CNN is evaluated in terms of compression ratio and coding efficiency compared to the original CNN. Evaluation results show that CNN could be compressed to about quarter with negligible coding loss by applying simple quantization to the trained weights.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.27 no.5
• /
• pp.690-698
• /
• 1990

Among the target identification algorithms presented up to now, the NNR algorithm utilize multifrequency and single observation method. But it is generally known that its implementation is extremely difficult. In this paper a new identification algorithm using the single frequency and cumulative observation method was proposed. The simulation result shows that the proposed algorithm with 31 observation and accumulation is more effective and realizable than the NNR with 3 freqnencies when the noise level n is below 0.3.

• Korean Journal of Plant Resources
• /
• v.21 no.1
• /
• pp.96-102
• /
• 2008

Nelumbo nucifera Gaertn.(family Nymphaeaceae) has been used for summer heat syndrome as home remedy in Japan, China and Korea. Although whole plant parts are edible, root is commonly consumed. It has been reported that rhizome extract showed anti-diabetic and anti-inflammatory effects. However, in spite of usefulness for treatment of various diseases, the effect of Nelumbo nucifera rhizome(NNR) on proliferation, migration and matrix degrading enzymes-matrix metalloproteinsases(MMPs), the expression of which degrades extracellular matrix(ECM) leading to metastasis, has not been fully elucidated. We examined the effect of hot water extract of NNR on the proliferation, migration and secretion of MMP-2 and MMP-9 in rat smooth muscle cells(rSMC), epidermoid cancer cells(A431) and breast cancer cells(MDA-MB-231). The proliferation assay was carried out using MTT assay, the principle of which depends upon the conversion of MTT by mitochondrial dehydrogensases of viable cells to formazan crystals. The effect of NNR on migration of cells was examined using wound healing assay. Our results showed that there was inhibition in the proliferation, migration and expression of MMP-2 and MMP-9 in dose dependent fashion in all the cells used. Thus, we concluded that NNR could be used as traditional medicine in the treatment of various diseases where proliferation, migration and MMPs' expression plays a pathological role like in restenosis and metastasis.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.6
• /
• pp.43-49
• /
• 2004

The classification time required by conventional multi-class SVMs(Support Vector Machines) greatly increases as the number of pattern classes increases. This is due to the fact that the needed set of binary class SVMs gets quite large. In this paper, we propose a method to reduce the number of classes by using nearest neighbor rule (NNR) in the principle component analysis and linear discriminant analysis (PCA+LDA) feature subspace. The proposed method reduces the number of face classes by selecting a few classes closest to the test data projected in the PCA+LDA feature subspace. Results of experiment show that our proposed method has a lower error rate than nearest neighbor classification (NNC) method. Though our error rate is comparable to the conventional multi-class SVMs, the classification process of our method is much faster.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.06a
• /
• pp.821-824
• /
• 2022

최근 딥러닝(Deep Learning) 기술이 다양한 분야에서 활용되고 있으며, 사전 학습된 딥러닝 모델에 대한 압축과 전송 방안에 관한 연구 또한 활발히 진행되고 있다. 이와 관련하여, 국제 표준화 기구인 ISO/IEC 산하 MPEG(Moving Picture Expert Group)에서는 인공신경망 모델을 다양한 딥러닝 프레임워크(Deep Learning Framework)에서 상호운용 가능한 포맷으로 압축 표현할 수 있는 NNC(Compression of Neural Network for Multimedia Content Description nd Analysis) 표준화를 진행하고 있다. 압축된 딥러닝 모델의 데이터를 효과적으로 저장하여 전송 및 사용하기 위해서는 ISOBMFF(ISO based Media File Format) 캡슐화 과정이 필요하다. 본 논문에서는 MPEG의 NNC 표준에 따라 사전 학습된 딥러닝 모델을 압축한 후 이를 통해 생성된 비트스트림(bitstream)을 ISOBMFF로 캡슐화하기 위한 기술을 제안 및 실험한다. 또한, 실험에 대한 검증을 위하여 생성된 ISOBMFF 데이터를 비트스트림으로 복원한 뒤 복호화하여 입력 비트스트림과 차이가 없음을 확인한다.

• The Korea Journal of Herbology
• /
• v.20 no.3
• /
• pp.75-81
• /
• 2005

Objectives : In brain disorders such as ischemic stroke, the final outcome depends largely on the duration and the degree of the ischemia as well as the susceptibility of various cell types in the affected brain region. In the present study, the effects of Nodus Nelumbinis Rhizomatis Extract(NNRe) were tested for the anti-oxidative action of rCBF. Methods : Regional cerebral blood flow(rCBF) were determined by LDF methods. LDF allows for real time, noninvasive, continuous recordings of local CBF. The LDF method has been widely used to trace hemodynamic changes in the superficial or the deep brain structures in experimental stroke research. Results : NNRe treatment showed no change on rCBF in methylene blue, ODQ and L-NNA pretreated rats. 120 minutes of MCAO and followed reperfusion, 0.1% concentration of NNR treatment improved the altered cerebral hemodynamics of cerebral ischemic by increasing rCBF. Conclusions : The ischemia/reperfusion induced oxidative stress may have contributed to cerebral damage in rats, and the present study provides clear evidences for the beneficial effect of NNR on ischemia/reperfusion induced brain injury.

• Proceedings of the PSK Conference
• /
• 2002.10a
• /
• pp.399.1-399.1
• /
• 2002

The chiral stationary phase derived from (＋) (18-crown-6)-2.3, 11.12-tetracarboxylic acid (18-C-6- TA) as a chiral selector has been employed for resolution of several $\alpha$-amino acids in HPLC. In a quest for the origin of chiral recognition of $\alpha$-amino acids in the presence of 18-C-6- T, A, as a chiral selector, these interactions responsible for the differential affinities shown toward enantioners were investigated by NNR spectroscopy. (omitted)