misc

Augusto links

https://www.programiz.com/python-programming/methods/built-in/zip https://stackoverflow.com/questions/39614516/using-a-numpy-array-to-assign-values-to-another-array https://stackoverflow.com/questions/41900738/assign-values-to-different-index-positions-in-numpy-array

Links

https://cdan.dot.gov/query (for reports National Highway Traffic Safety Administration - USA) https://rspcb.safety.fhwa.dot.gov/Dashboard/Default.aspx

Metric Learning & Learning things

Focal Loss - https://arxiv.org/pdf/1708.02002.pdf Focal Loss implementation https://kornia.readthedocs.io/en/latest/losses.html#kornia.losses.FocalLoss https://researchweb.iiit.ac.in/~praveen.krishnan/compreExam/slides/T4-DistanceMetricLearningBeyond0-1Loss.pdf https://arxiv.org/pdf/1708.01682.pdf https://arxiv.org/abs/1812.05944 http://researchers.lille.inria.fr/abellet/talks/metric_learning_tutorial_CIL.pdf https://github.com/KevinMusgrave/pytorch-metric-learning http://www.bmva.org/bmvc/2016/papers/paper119/paper119.pdf

http://contrib.scikit-learn.org/metric-learn/introduction.html

what if we add the xxxx% of variance as (1-xxx) or something similar into the loss? the fact is that I want to nt have all the 512 elementes meaning something, but a small representation.

https://www.powerthesaurus.org/other/synonyms https://arxiv.org/pdf/2004.05937.pdf https://towardsdatascience.com/knowledge-distillation-simplified-dd4973dbc764 https://arxiv.org/pdf/1912.13179.pdf

https://www.quora.com/What-is-a-teacher-student-model-in-a-Convolutional-neural-network

http://www.bmva.org/bmvc/2016/papers/paper119/paper119.pdf https://arxiv.org/pdf/1708.01682.pdf https://ieeexplore.ieee.org/document/1041355 https://ieeexplore.ieee.org/document/5700786 https://ieeexplore.ieee.org/document/6005622 https://ieeexplore.ieee.org/document/7727849