nn-utils 0.1.0

Useful machine-learning related stuff.

Typical usage often looks like this:

#!/usr/bin/env python
import ml_utils as mlu
mlu.imutils.resize_keep_aspect(img_arr):

Some functions for supporting machine learning, generally and caffe-specific. Currently just a dump, ideally will clean up and add install.py, examples etc

caffe - python layers for

• image read (file or lmdb) - single / multi label, pixel level labels

• on-the-fly augmentation

• support for pixel-level segmentation (read mask as label, etc)

• controlling and reporting acc/loss of solver with solver.step

image processing

• image augmentation - including bounding boxes and pixel level

• acccuracy/precision/recall reporting for single label, multilabel , bounding box, and pixel level

• utilities e.g. read from anywhere (url/db/local file/img array)

• grabcut

read/write various file formats

• lmdb, hd5

• yolo

• deepfashion

• tamara berg

• ILSRVC

• etc

tagging tools

• multilabel

• pixel level