Skip to content

Latest commit

 

History

History
121 lines (88 loc) · 5.41 KB

OBJECT_DETECTION_NOTES.md

File metadata and controls

121 lines (88 loc) · 5.41 KB

Object detection notes

Courses, tutorials

Image processing libraries

Libraries for deep learning

Models

Data formats

Labelling tools

Commercial tools

  • Roboflow makes the conversion between formats.

Pattern Analysis, Statistical Modelling and computational Learning

Image augmentation.

Trainin pytorch, maskrcnn

Format conversion

Pascal VOC XML

YOLOv4 needs YOLOv4 Pytorch TXT

Source codes formats

Which data is necessary for PASCAL VOC? Read and Write Pascal Voc XML Annotations in Python VIA2COCO via2coco pycococreator

https://patrickwasp.com/create-your-own-coco-style-dataset/ How to Automatically Generate VGG Image Annotation Files.

TODO to clean

https://github.com/michhar/pytorch-mask-rcnn-samples https://www.kaggle.com/code/pestipeti/pytorch-starter-fasterrcnn-train https://www.kaggle.com/code/pestipeti/pytorch-starter-fasterrcnn-train https://christianjmills.com/posts/icevision-mask-rcnn-tutorial/part-1/ https://github.com/michhar/pytorch-mask-rcnn-samples/blob/master/Train.ipynb https://pytorch.org/vision/stable/models.html#object-detection-instance-segmentation-and-person-keypoint-detection https://haochen23.github.io/2020/06/fine-tune-mask-rcnn-pytorch.html#.ZBCQux_MJPY https://towardsdatascience.com/train-mask-rcnn-net-for-object-detection-in-60-lines-of-code-9b6bbff292c3 https://debuggercafe.com/instance-segmentation-with-pytorch-and-mask-r-cnn/ https://github.com/topics/annotation-conversion https://colab.research.google.com/drive/16jcaJoc6bCFAQ96jDe2HwtXj7BMD_-m5#scrollTo=PIbAM2pv-urF&line=4&uniqifier=1 https://www.kaggle.com/code/kerneler/starter-mask-rcnn-pytorch-weights-cb974d86-a https://towardsdatascience.com/train-mask-rcnn-net-for-object-detection-in-60-lines-of-code-9b6bbff292c3

Comenzar a indagar otro nuevo detector de objetos generales. Mirar el código fuente de Pytorch mobile Android. Probar el detector de personas en Windows. Subirlo a Github. Resolver el tema de la GPU. META: TENER UN DETECTOR DE OBJETOS. Necesito ver como enlazar un detector de objetos al formato que ya tengo. Colocar un detector de objetos varios para hacer andar con la cámara web. Ver los tutoriales: https://pytorch.org/blog/torchvision03/ https://medium.com/jumio/object-detection-tutorial-with-torchvision-82b8f269f6ff https://pytorch.org/tutorials/intermediate/torchvision_tutorial.html https://pytorch.org/vision/stable/models.html

Buen tutorial https://debuggercafe.com/object-detection-using-pytorch-faster-r-cnn-mobilenetv3/ https://colab.research.google.com/drive/1eAUjzV3nZXkUXi0spPg6zUHtJWaUSzFk?usp=sharing Good option to train https://github.com/sovit-123/fasterrcnn-pytorch-training-pipeline

https://pytorch.org/vision/stable/models.html#initializing-pre-trained-models https://pytorch.org/vision/stable/models.html#object-detection

https://github.com/pytorch/pytorch#a-gpu-ready-tensor-library