• 한국멀티미디어학회논문지
• /
• 제25권1호
• /
• pp.39-51
• /
• 2022

Image-based gender classification and age estimation of human are classic problems in computer vision. Most of researches in this field focus just only one task of either gender classification or age estimation and most of the reported methods for each task focus on accuracy performance and are not computationally light. Thus, running both tasks together simultaneously on low cost mobile or embedded systems with limited cpu processing speed and memory capacity are practically prohibited. In this paper, we propose a novel light-weight gender classification and age estimation method based on ensemble multitasking deep learning with light-weight processing neural network architecture, which processes both gender classification and age estimation simultaneously and in real-time even for embedded systems. Through experiments over various well-known datasets, it is shown that the proposed method performs comparably to the state-of-the-art gender classification and/or age estimation methods with respect to accuracy and runs fast enough (average 14fps) on a Jestson Nano embedded board.

• Genomics & Informatics
• /
• 제15권4호
• /
• pp.156-161
• /
• 2017

Transcriptome analysis has been widely used to make biomarker panels to diagnose cancers. In breast cancer, the age of the patient has been known to be associated with clinical features. As clinical transcriptome data have accumulated significantly, we classified all human genes based on age-specific differential expression between normal and breast cancer cells using public data. We retrieved the values for gene expression levels in breast cancer and matched normal cells from The Cancer Genome Atlas. We divided genes into two classes by paired t test without considering age in the first classification. We carried out a secondary classification of genes for each class into eight groups, based on the patterns of the p-values, which were calculated for each of the three age groups we defined. Through this two-step classification, gene expression was eventually grouped into 16 classes. We showed that this classification method could be applied to establish a more accurate prediction model to diagnose breast cancer by comparing the performance of prediction models with different combinations of genes. We expect that our scheme of classification could be used for other types of cancer data.

• Journal of Oral Medicine and Pain
• /
• 제24권1호
• /
• pp.95-106
• /
• 1999

This study was performed to investigate the age distribution with tooth calcification and degree of eruption of permanent teeth. For the study, healthy 184 patients from 5 to 19 years old without any previous serious dental treatment were randomly selected, and intraoral standard films and dental casts were taken for evaluation of stage of calcification and degree of eruption, respectively. Tooth calcification of 13 stages, designed by the author based on the Nolla's classification and eruption level of 4 or 5 degree was used. Data were processed by SAS/Stat program and the obtained results were as follows; 1. The age of root completed with open apex in lower posterior teeth were 13.8 years for first premolar, 14.0 years for second premolar, 10.5 years for first molar, and 14.2 years for second molar. There were no significant difference between right and left side. 2. As for the sequence of eruption, first molar was the first teeth erupted in upper arch, while central incisor was the first teeth in lower arch. In general, eruption of lower teeth were slightly earlier than the corresponding teeth of upper arch. 3. There were no difference of age of the same stage of development between Nolla's and the author's classification. From the results, the author's classification can be used for estimation of age with more finely in age of 8 to 15 years old. 4. Multiple regression equations for age with Nolla's(Ns) and the author's(Ks) classification of tooth calcification, and degree of eruption(DE) were as follow; Age(by #34) = 7.55 + 0.76Ks34 + 0.80DE34 - 0.72Ns34 Age(by #35) = 7.10 + 0.81Ks35 + 0.6IDE35 Age(by #37) = 6.61 + 0.82Ks37 + 0.5IDE37. Age(by #44) = 7.02 + 0.62Ks44 + 0.82DE44 Age(by #45) = 8.04 + 0.93Ks45 + 0.64DE45 - 0.89Ns45 Age(by #47) = 6.40 + 0.86Ks47 + 0.56DE47.

• 한국항행학회논문지
• /
• 제16권1호
• /
• pp.151-157
• /
• 2012

최근 영상처리기술과 컴퓨터과학의 발달로 연령변화에 따른 얼굴형상 분류 방법은 일반적인 주제가 되었다. 사람의 연령별 얼굴분류는 생물학적 유전자와 오랜 생활의 식습관으로 인하여 얼굴 형상이 변하기 때문에 통계적 형상만으로 예측하기란 쉽지 않다. 본 논문에서는 Gobor 특징과 fuzzy SVM 기법을 이용하여 연령대별 얼굴분류 기법을 제안하였다. Gabor 웨이블릿 함수는 얼굴의 특징벡터를 구하기 위하여 사용되고 연령대별 얼굴형상 구분이 애매모호한 문제를 해결하기 위해 fuzzy SVM 기법을 이용하여 연령별 소속 함수를 정의하였다. 제안한 방법으로 연령별 소속함수에 따른 얼굴 분류 실험을 수행하였고 제안한 방법의 타당성을 확인하였다.

• 한국전자통신학회논문지
• /
• 제17권1호
• /
• pp.99-104
• /
• 2022

인공지능은 놀라운 이점으로 우리 삶의 중요한 부분을 차지하고 있다. 기계는 이미지에서 물체를 인식하는 것, 특히 사람들을 정확한 나이와 성별 그룹으로 분류하는 것에 있어서 인간을 능가하고 있다. 이러한 측면에서 나이와 성별 분류는 최근 수십 년 동안 컴퓨터 비전 연구자들 사이에서 뜨거운 주제 중 하나였다. 심층 합성곱 신경망(CNN) 모델의 배포는 최첨단 성능을 달성했다. 그러나 대부분의 CNN 기반 아키텍처는 수십 개의 훈련 매개 변수로 매우 복잡하기 때문에 많은 계산 시간과 자원이 필요하다. 이러한 이유로 기존 방법에 비해 훈련 매개 변수와 훈련 시간이 현저히 적은 새로운 CNN기반 분류 알고리즘을 제안한다. 덜 복잡함에도 불구하고 우리 모델은 UTKFace 데이터 세트에서 연령 및 성별 분류의 더 나은 정확도를 보여준다.

• 아동학회지
• /
• 제20권3호
• /
• pp.259-273
• /
• 1999

This studied investigated children's music identification, classification, and seriation cognitive task performance abilities by age and sex. The subjects were l20 six-, eight-, and ten-year-old school children. There were significant positive correlations among music cognition tasks and significant age and sex differences within each of the music tasks. Ten-year-old children were more likely to complete their music identification tasks than the younger children and girls were more likely than boys to complete their music identification tasks. Eight- and 10-year-old children were more likely to complete their music classification tasks than the younger group. Piagetian stage theory was demonstrated in children's music classification task performance. There was an age-related increase in the performance of the music seriation tasks. Developmental sequential theory was demonstrated in music seriation performance.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2023년도 추계학술발표대회
• /
• pp.568-570
• /
• 2023

The problem of classifying of age, gender, and race images still poses challenges. Despite deep and machine learning strides, convolutional neural networks (CNNs) remain pivotal in addressing these issues. This paper introduces a novel CNN-based approach for accurate and efficient age, gender, and race classification. Leveraging CNNs with residual blocks, our method enhances learning while minimizing computational complexity. The model effectively captures low-level and high-level features, yielding improved classification accuracy. Evaluation of the diverse 'fair face' dataset shows our model achieving 56.3%, 94.6%, and 58.4% accuracy for age, gender, and race, respectively.

• 전자공학회논문지
• /
• 제51권2호
• /
• pp.69-75
• /
• 2014

기존의 피플 카운팅 시스템(People Counting System)은 주로 오버헤드(Overhead) 시점에 설치된 단일 카메라를 활용하기 때문에, 겹침 및 가림 현상과 일정 크기 이상의 소지품, 급격한 조명 변화와 같은 외부 환경적 요인들로 인해 객체 인식에 장애가 발생하고, 다양한 환경에서 카운팅을 수행하기에 어려움이 존재한다. 이에 본 논문에서는 기존 단일 시점 피플 카운팅 시스템의 인식 장애 개선 및 외부 환경적 요인들에 보다 강인하게 카운팅할 수 있도록, 오버헤드 및 전면 시점에 두 개의 깊이 카메라를 활용하는 얼굴 나이 그룹별 피플 카운팅 시스템을 제안한다. 제안 시스템은 오버헤드 영상 처리, 전면 영상 처리, 동일 객체 판별, 얼굴 나이 그룹 분류, 입퇴장 카운팅의 총 5가지 처리를 통해 얼굴 나이 그룹별 피플 카운팅을 수행한다. 제안 시스템을 C++, OpenCV 및 Kinect SDK를 기반으로 구현하여, 나이 그룹별로 10명씩 총 40명을 대상으로 피플 카운팅 성능과 나이 그룹 분류 성능을 각각 평가하였다. 성능 평가 결과는 피플 카운팅에서 약 98%의 정확도를 나타냈고, 나이 그룹 분류는 약 74.23%의 정확도를 보였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제8권7호
• /
• pp.2480-2495
• /
• 2014

This paper describes a hierarchical method for image-based age estimation that combines age group classification and age value estimation. The proposed method uses a coarse-to-fine strategy with different appearance features to describe facial shape and texture. Considering the damage to continuity between neighboring groups caused by fixed divisions during age group classification, a dynamic grouping technique is employed to allow non-fixed groups. Based on the given group, an ordinal hyperplane ranking (OHRank) model is employed to transform age estimation into a series of binary enquiry problems that can take advantage of the intrinsic correlation and ordinal information of age. A set of experiments on FG-NET are presented and the results demonstrate the validity of our solution.

• 대한족부족관절학회지
• /
• 제10권2호
• /
• pp.196-200
• /
• 2006

Purpose: To study prognostic factors of unilateral calcaneus fracture underwent surgery. Materials and Methods: We selected appropriate 60 cases of 236 calcaneus fracture cases between March 1985 and March 2004, and analyzed the correlation between sex, age, smoking, injury mechanism, Essex-Lopresti classification of calcaneus fracture, preoperative Bohler angle, postoperative Bohler angle, postoperative 1 year Bohler angle and Visual Analogue Scale (VAS), P.S. Kerr's Calcaneal Fracture Score (CFSS). The average age was 41.4 and average follow up period was 74 (12 to 240) months. Results: For follow up period, average VAS is 3.43 and CFSS is 81.23. The sex, age, smoking, injury mechanism, and preoperative, postoperative, postoperative 1 year Bohler angle had no correlation with the prognosis. But the Essex-Lopresti classification of calcaneus fracture, tongue type had better prognosis than joint depression type (VAS : p=0.041, CFSS : p=0.021). Conclusion: In unilateral calcaneus fracture, the sex, age, smoking, injury mechanism, preoperative Bohler angle, postoperative Bohler angle, postoperative 1 year Bohler angle had no correlation with the prognosis of fracture, but in Essex-Lopresti classification, the tongue type fracture had better prognosis than the joint depression type.