Arc Lab.'s Blog

[업데이트 2018.09.24 16:41] 1. 베이스 논문 [논문 요약 40] Traffic-sign detection and classification in the wild (http://arclab.tistory.com/205) 2. 베이스 논문을 인용한 논문 Detecting Small Signs from Large Images (https://ieeexplore.ieee.org/abstract/document/8102940) 3. 주요 내용 요약 3.1. 연구 내용 파악- 무엇에 관한 연구인가?Computer vision분야의 object detection에 대한 연구이며, 교통 표지판과 같은 매우 작은 크기의 사물 인식을 주제로 작성된 논문입니다. 본 논문에서 제안한 patch-level의 o..

[업데이트 2018.11.02 15:44] 사십번째 요약할 논문은 "Traffic-sign detection and classification in the wild" (https://www.cv-foundation.org/openaccess/content_cvpr_2016/papers/Zhu_Traffic-Sign_Detection_and_CVPR_2016_paper.pdf) 입니다. 인트로에 대해 요약한 내용은 아래와 같습니다. 기존의 데이터셋의 경우 실제 환경에서 인식하려는 대상 사물의 사이즈가 아주 작은 경우에 대해 학습 및 테스트하기 적합하지 않은데, 본 논문에서는 실제 환경과 유사한 데이터셋을 제공합니다. * Benchmark 테스트를 위해 환경 설정 및 학습, 평가 등에 대해 GitHub에 정..

[업데이트 2018.09.24 14:29] 1. 베이스 논문[논문 요약17] Perceptual Generative Adversarial Networks for Small Object Detection(http://arclab.tistory.com/172) 2. 레퍼런스 논문[23] H. Li, Z. Lin, X. Shen, J. Brandt, and G. Hua. A convolutional neural network cascade for face detection. In CVPR, pages 5325–5334, 2015. 1(http://users.eecs.northwestern.edu/~xsh835/assets/cvpr2015_cascnn.pdf) Some efforts [4, 25, 18, 39,..

[업데이트 2018.09.24 14:22] 1. 베이스 논문[논문 요약17] Perceptual Generative Adversarial Networks for Small Object Detection(http://arclab.tistory.com/172) 2. 레퍼런스 논문[22] C. Li and M. Wand. Combining markov random fields and con- volutional neural networks for image synthesis. arXiv preprint arXiv:1601.04589, 2016. 2(https://arxiv.org/pdf/1601.04589.pdf) In [22] and [40], GANs were employed to learn a mappi..

[업데이트 2018.09.24 14:21] 1. 베이스 논문[논문 요약17] Perceptual Generative Adversarial Networks for Small Object Detection(http://arclab.tistory.com/172) GANs were also applied to image super- resolution in [21]. 2. 레퍼런스 논문[21] C.Ledig,L.Theis,F.Husza ́r,J.Caballero,A.Aitken,A.Tejani, J. Totz, Z. Wang, and W. Shi. Photo-realistic single image super- resolution using a generative adversarial network. arXiv..

[업데이트 2018.09.08 17:41] 1. 베이스 논문[논문 요약17] Perceptual Generative Adversarial Networks for Small Object Detection(http://arclab.tistory.com/172) Traffic sign detection and recognition has been a popular problem in intelligent vehicles, and various methods [20, 15, 34, 19, 38, 45] have been pro- posed to address this challenging task. Traditional methods for this task includes [20] [15]. 2. 레퍼런스 논..

[업데이트 2018.09.08 17:19] 1. 베이스 논문[논문 요약17] Perceptual Generative Adversarial Networks for Small Object Detection(http://arclab.tistory.com/172) Jin et al. [19] proposed to train the CNN with hingle loss, which provides better test accuracy and faster stable convergence. 2. 레퍼런스 논문[19] J. Jin, K. Fu, and C. Zhang. Traffic sign recognition with hinge loss trained convolutional neural networks. IEE..

[업데이트 2018.09.08 17:00] 1. 베이스 논문[논문 요약17] Perceptual Generative Adversarial Networks for Small Object Detection(http://arclab.tistory.com/172) Some efforts [4, 25, 18, 39, 23, 1] have been devoted to addressing small object detection problems. One common practice [4, 25] is to increase the scale of input images to enhance the resolution of small objects and produce high-resolution feature maps...

[업데이트 2018.09.08 16:28] 1. 베이스 논문[논문 요약17] Perceptual Generative Adversarial Networks for Small Object Detection(http://arclab.tistory.com/172) The implementation is based on the publicly available Fast R-CNN framework [11] built on the Caffe platform [17]. 2. 레퍼런스 논문[17] Y. Jia, E. Shelhamer, J. Donahue, S. Karayev, J. Long, R. Girshick, S. Guadarrama, and T. Darrell. Caffe: Convolutional archi..

[업데이트 2018.09.08 16:28] 1. 베이스 논문[논문 요약17] Perceptual Generative Adversarial Networks for Small Object Detection(http://arclab.tistory.com/172) Following [16], we perform down-sampling directly by convolutional layers with a stride of 2. 2. 레퍼런스 논문[16] K. He, X. Zhang, S. Ren, and J. Sun. Deep residual learning for image recognition. arXiv preprint arXiv:1512.03385, 2015. 5(https://arxiv.org/pdf..