• ETRI Journal
• /
• v.42 no.2
• /
• pp.230-238
• /
• 2020

Human-object interaction (HOI) detection is a popular computer vision task that detects interactions between humans and objects. This task can be useful in many applications that require a deeper understanding of semantic scenes. Current HOI detection networks typically consist of a feature extractor followed by detection layers comprising small filters (eg, 1 × 1 or 3 × 3). Although small filters can capture local spatial features with a few parameters, they fail to capture larger context information relevant for recognizing interactions between humans and distant objects owing to their small receptive regions. Hence, we herein propose a three-stream HOI detection network that employs a context convolution module (CCM) in each stream branch. The CCM can capture larger contexts from input feature maps by adopting combinations of large separable convolution layers and residual-based convolution layers without increasing the number of parameters by using fewer large separable filters. We evaluate our HOI detection method using two benchmark datasets, V-COCO and HICO-DET, and demonstrate its state-of-the-art performance.

• Resources Recycling
• /
• v.9 no.6
• /
• pp.9-14
• /
• 2000

Comparing the adsorption characteristics of coconut shell activated carbon (CSAC) and waste paint activated carbon (WPAC), Freundlich adsorption isotherms of alkylbenzene sulfonate (ABS) obtained from the secondary treatment water of H company and effluent of D company were estimated q=23.12 $C^{0.42}$ , q=18.32 $C^{0.38}$ with WPAC and $q=36.76C^{1.37}$ /, q=26.67 $C^{0.42}$ with CSAC respectively. In the case of H company, breakthrough time of the ABS using CSAC by continuous experiment was estimated 680 minute md that of WPAC was 610 minute. In the case of D company effluent, CSAC was estimated 720 minute, and that of WPAC was estimated 640 minute to reach the breakthrough. From the above results, it is possible to replace the coco-nut shell activated carbon with wasted paint activated carbon.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.10
• /
• pp.3608-3626
• /
• 2021

The detection of bolts is an important task in high-speed train inspection systems, and it is frequently performed to ensure the safety of trains. The difficulty of the vision-based bolt inspection system lies in small sample defect detection, which makes the end-to-end network ineffective. In this paper, the problem is resolved in two stages, which includes the detection network and cascaded classification networks. For small bolt detection, all bolts including defective bolts and normal bolts are put together for conducting annotation training, a new loss function and a new boundingbox selection based on the smallest axis-aligned convex set are proposed. These allow YOLOv3 network to obtain the accurate position and bounding box of the various bolts. The average precision has been greatly improved on PASCAL VOC, MS COCO and actual data set. After that, the Siamese network is employed for estimating the status of the bolts. Using the convolutional Siamese network, we are able to get strong results on few-shot classification. Extensive experiments and comparisons on actual data set show that the system outperforms state-of-the-art algorithms in bolt inspection.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.9
• /
• pp.4665-4683
• /
• 2019

Previous methods build image annotation model by leveraging three basic dependencies: relations between image and label (image/label), between images (image/image) and between labels (label/label). Even though plenty of researches show that multiple dependencies can work jointly to improve annotation performance, different dependencies actually do not "work jointly" in their diagram, whose performance is largely depending on the result predicted by image/label section. To address this problem, we propose the adaptive attention annotation model (AAAM) to associate these dependencies with the prediction path, which is composed of a series of labels (tags) in the order they are detected. In particular, we optimize the prediction path by detecting the relevant labels from the easy-to-detect to the hard-to-detect, which are found using Binary Cross-Entropy (BCE) and Triplet Margin (TM) losses, respectively. Besides, in order to capture the inforamtion of each label, instead of explicitly extracting regional featutres, we propose the self-attention machanism to implicitly enhance the relevant region and restrain those irrelevant. To validate the effective of the model, we conduct experiments on three well-known public datasets, COCO 2014, IAPR TC-12 and NUSWIDE, and achieve better performance than the state-of-the-art methods.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.12
• /
• pp.4816-4834
• /
• 2020

This paper proposes transfer learning and fine-tuning techniques for a deep learning model to detect three distinct brain tumors from Magnetic Resonance Imaging (MRI) scans. In this work, the recent YOLOv4 model trained using a collection of 3064 T1-weighted Contrast-Enhanced (CE)-MRI scans that were pre-processed and labeled for the task. This work trained with the partial 29-layer YOLOv4-Tiny and fine-tuned to work optimally and run efficiently in most platforms with reliable performance. With the help of transfer learning, the model had initial leverage to train faster with pre-trained weights from the COCO dataset, generating a robust set of features required for brain tumor detection. The results yielded the highest mean average precision of 93.14%, a 90.34% precision, 88.58% recall, and 89.45% F1-Score outperforming other previous versions of the YOLO detection models and other studies that used bounding box detections for the same task like Faster R-CNN. As concluded, the YOLOv4-Tiny can work efficiently to detect brain tumors automatically at a rapid phase with the help of proper fine-tuning and transfer learning. This work contributes mainly to assist medical experts in the diagnostic process of brain tumors.

• Korean Journal of Remote Sensing
• /
• v.37 no.4
• /
• pp.747-761
• /
• 2021

With the increasing application of UAVs in intelligent transportation systems, vehicle detection for aerial images has become an essential engineering technology and has academic research significance. In this paper, a vehicle detection method for aerial images based on the YOLOv4 deep learning algorithm is presented. At present, the most known datasets are VOC (The PASCAL Visual Object Classes Challenge), ImageNet, and COCO (Microsoft Common Objects in Context), which comply with the vehicle detection from UAV. An integrated dataset not only reflects its quantity and photo quality but also its diversity which affects the detection accuracy. The method integrates three public aerial image datasets VAID, UAVD, DOTA suitable for YOLOv4. The training model presents good test results especially for small objects, rotating objects, as well as compact and dense objects, and meets the real-time detection requirements. For future work, we will integrate one more aerial image dataset acquired by our lab to increase the number and diversity of training samples, at the same time, while meeting the real-time requirements.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.7
• /
• pp.2328-2344
• /
• 2022

2D human pose estimation still faces difficulty in low-resolution images. Most existing top-down approaches scale up the target human bonding box images to the large size and insert the scaled image into the network. Due to up-sampling, artifacts occur in the low-resolution target images, and the degraded images adversely affect the accurate estimation of the joint positions. To address this issue, we propose a multi-resolution input feature fusion network for human pose estimation. Specifically, the bounding box image of the target human is rescaled to multiple input images of various sizes, and the features extracted from the multiple images are fused in the network. Moreover, we introduce a guiding channel which induces the multi-resolution input features to alternatively affect the network according to the resolution of the target image. We conduct experiments on MS COCO dataset which is a representative dataset for 2D human pose estimation, where our method achieves superior performance compared to the strong baseline HRNet and the previous state-of-the-art methods.

• Journal of Plant Biotechnology
• /
• v.47 no.4
• /
• pp.330-336
• /
• 2020

Ex-situ conservation of the ornamental and medicinal orchid, Coelogyne stricta, was performed by mass propagation using seed culture. Propagation stages were optimized using full- and half-strength solidified MS medium with different phytohormones. Maximum seed germination (88 ± 0.5% over 6 weeks of culture) was achieved on half-strength MS medium supplemented with 15% coconut water. Maximum shoot numbers were found on full-strength MS medium supplemented with 1 mg/L BAP, 2 mg/L Kinetin, and 10% coconut water, while the longest root was developed on full-strength MS medium with 1.5 mg/L IBA. A 2:1:1 combination of coco-peat, pine bark, and sphagnum moss was found to be a suitable potting mixture resulting in 80% seedling survivability. The cytotoxic activity of extracts of both wild plants and in vitro-developed protocorms was determined using an MTT (3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide) assay on a cervical cancer cell line. The wild plant extract inhibited the growth of 41.99% of cells, showing that this extract has moderate cytotoxic activity toward cervical cancer cells.

• Journal of Microbiology and Biotechnology
• /
• v.31 no.3
• /
• pp.429-438
• /
• 2021

Bacterial cellulose (BC) is widely used in the food industry for products such as nata de coco. The mechanical properties of BC hydrogels, including stiffness and viscoelasticity, are determined by the hydrated fibril network. Generally, Komagataeibacter bacteria produce gluconic acids in a glucose medium, which may affect the pH, structure and mechanical properties of BC. In this work, the effect of pH buffer on the yields of Komagataeibacter hansenii strain ATCC 53582 was studied. The bacterium in a phosphate and phthalate buffer with low ionic strength produced a good BC yield (5.16 and 4.63 g/l respectively), but there was a substantial reduction in pH due to the accumulation of gluconic acid. However, the addition of gluconic acid enhanced the polymer density and mechanical properties of BC hydrogels. The effect was similar to that of the bacteria using glycerol in another carbon metabolism circuit, which provided good pH stability and a higher conversion rate of carbon. This study may broaden the understanding of how carbon sources affect BC biosynthesis.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.11a
• /
• pp.258-261
• /
• 2022

최근 딥러닝 기술이 발전하면서 이미지를 설명하는 캡션을 생성하는 모델 또한 발전하였다. 하지만 기존 이미지 캡션 모델은 대다수 영어로 구현되어있어 영어로 캡션을 생성하게 된다. 따라서 한국어 캡션을 생성하기 위해서는 영어 이미지 캡션 결과를 한국어로 번역하는 과정이 필요하다는 문제가 있다. 이에 본 연구에서는 기존의 이미지 캡션 모델을 이용하여 한국어 캡션을 직접 생성하는 모델을 만들고자 한다. 이를 위해 이미지 캡션 모델 중 잘 알려진 Show, Attend and Tell 모델을 이용하였다. 학습에는 MS-COCO 데이터의 한국어 캡션 데이터셋을 이용하였다. 한국어 형태소 분석기를 이용하여 토큰을 만들고 캡션 모델을 재학습하여 한국어 캡션을 생성할 수 있었다. 만들어진 한국어 이미지 캡션 모델은 BLEU 스코어를 사용하여 평가하였다. 이때 BLEU 스코어를 사용하여 생성된 한국어 캡션과 영어 캡션의 성능을 평가함에 있어서 언어의 차이에 인한 결과 차이가 발생할 수 있으므로, 영어 이미지 캡션 생성 모델의 출력을 한국어로 번역하여 같은 언어로 모델을 평가한 후 최종 성능을 비교하였다. 평가 결과 한국어 이미지 캡션 생성 모델이 영어 이미지 캡션 생성 모델을 한국어로 번역한 결과보다 좋은 BLEU 스코어를 갖는 것을 확인할 수 있었다.