EfficientNet: 학습 속도 테스트
https://pizpaz.github.io/paper/ml/EfficientNet-Rethinking-Model-Scaling-for-Convolutional-Neural-Networks 에서 소개했던 Efficientnet의 학습 속도를 테스트해봤다.
Posts by Category
ML
EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
https://arxiv.org/abs/1905.11946
Tensorflow Multi-machine learning test
회사에서 학습셋으로 몇천만장 단위의 데이터를 뽑을 수 있게 되었는데 100만장 정도의 학습셋을 학습한 경험 밖에 없다. 이 상태에선 파라미터 튜닝은 기다리다 지쳐 못한다.
Ranked List Loss for Deep Metric Learning
https://arxiv.org/abs/1903.03238, CVPR2019
Visual Search At Alibaba
https://102.alibaba.com/downloadFile.do?file=1534297833520/VisualSearch.pdf
DropBlock: A regularization method for convolutional networks
https://arxiv.org/abs/1810.12890, Google Brain
Don’t Decay the Learning Rate, Increase the Batch Size
https://arxiv.org/pdf/1711.00489.pdf, ICLR2018, Google Brain.
Wide-Slice Residual Networks for Food Recognition
https://arxiv.org/abs/1612.06543
Menu-Match: Restaurant-Specific Food Logging from Images
https://ieeexplore.ieee.org/document/7045971, Microsoft, 2015
Harmonic Networks: Deep Translation and Rotation Equivariance
https://arxiv.org/abs/1612.04642
FooDD: Food Detection Dataset for Calorie Measurement Using Food Images
http://www.site.uottawa.ca/%7Eshervin/pubs/FoodRecognitionDataset-MadiMa.pdf
ChineseFoodNet: A Large-scale Image Dataset for Chinese Food Recognition
https://arxiv.org/abs/1705.02743
Exploring the Limits of Weakly Supervised Pretraining
https://research.fb.com/wp-content/uploads/2018/05/exploring_the_limits_of_weakly_supervised_pretraining.pdf, Facebook
Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition
https://arxiv.org/pdf/1406.4729v1.pdf, ECCV2014
Learning a Discriminative Filter Bank within a CNN for Fine-grained Recognition
http://openaccess.thecvf.com/content_cvpr_2018/papers/Wang_Learning_a_Discriminative_CVPR_2018_paper.pdf, CVPR2018
Deep Learning is Robust to Massive Label Noise
https://openreview.net/pdf?id=B1p461b0W
A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise
http://citeseerx.ist.psu.edu/viewdoc/download;jsessionid=FBB2448080DA6C55479F67DE02132AF9?doi=10.1.1.71.1980&rep=rep1&type=pdf
Improved Deep Metric Learning with Multi-class N-pair Loss Objective
http://www.nec-labs.com/uploads/images/Department-Images/MediaAnalytics/papers/nips16_npairmetriclearning.pdf
paper
EfficientNet: 학습 속도 테스트
https://pizpaz.github.io/paper/ml/EfficientNet-Rethinking-Model-Scaling-for-Convolutional-Neural-Networks 에서 소개했던 Efficientnet의 학습 속도를 테스트해봤다.
EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
https://arxiv.org/abs/1905.11946
Ranked List Loss for Deep Metric Learning
https://arxiv.org/abs/1903.03238, CVPR2019
Visual Search At Alibaba
https://102.alibaba.com/downloadFile.do?file=1534297833520/VisualSearch.pdf
DropBlock: A regularization method for convolutional networks
https://arxiv.org/abs/1810.12890, Google Brain
Don’t Decay the Learning Rate, Increase the Batch Size
https://arxiv.org/pdf/1711.00489.pdf, ICLR2018, Google Brain.
Wide-Slice Residual Networks for Food Recognition
https://arxiv.org/abs/1612.06543
Menu-Match: Restaurant-Specific Food Logging from Images
https://ieeexplore.ieee.org/document/7045971, Microsoft, 2015
Harmonic Networks: Deep Translation and Rotation Equivariance
https://arxiv.org/abs/1612.04642
FooDD: Food Detection Dataset for Calorie Measurement Using Food Images
http://www.site.uottawa.ca/%7Eshervin/pubs/FoodRecognitionDataset-MadiMa.pdf
ChineseFoodNet: A Large-scale Image Dataset for Chinese Food Recognition
https://arxiv.org/abs/1705.02743
Exploring the Limits of Weakly Supervised Pretraining
https://research.fb.com/wp-content/uploads/2018/05/exploring_the_limits_of_weakly_supervised_pretraining.pdf, Facebook
Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition
https://arxiv.org/pdf/1406.4729v1.pdf, ECCV2014
Learning a Discriminative Filter Bank within a CNN for Fine-grained Recognition
http://openaccess.thecvf.com/content_cvpr_2018/papers/Wang_Learning_a_Discriminative_CVPR_2018_paper.pdf, CVPR2018
Deep Learning is Robust to Massive Label Noise
https://openreview.net/pdf?id=B1p461b0W
A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise
http://citeseerx.ist.psu.edu/viewdoc/download;jsessionid=FBB2448080DA6C55479F67DE02132AF9?doi=10.1.1.71.1980&rep=rep1&type=pdf
Improved Deep Metric Learning with Multi-class N-pair Loss Objective
http://www.nec-labs.com/uploads/images/Department-Images/MediaAnalytics/papers/nips16_npairmetriclearning.pdf
Tensorflow
EfficientNet: 학습 속도 테스트
https://pizpaz.github.io/paper/ml/EfficientNet-Rethinking-Model-Scaling-for-Convolutional-Neural-Networks 에서 소개했던 Efficientnet의 학습 속도를 테스트해봤다.
Tensorflow Multi-machine learning test
회사에서 학습셋으로 몇천만장 단위의 데이터를 뽑을 수 있게 되었는데 100만장 정도의 학습셋을 학습한 경험 밖에 없다. 이 상태에선 파라미터 튜닝은 기다리다 지쳐 못한다.