Skip to content

Latest commit

 

History

History
72 lines (48 loc) · 2.96 KB

README.md

File metadata and controls

72 lines (48 loc) · 2.96 KB
Raw

text-recognition-resnet-fc

Use-case and high-level description

text-recognition-resnet-fc is a simple and preformant scene text recognition model based on ResNet with Fully Connected text recognition head. Source implementation on a PyTorch* framework could be found here. Model is able to recognize alphanumeric text.

Specification

Metric Value
Type Scene Text Recognition
GFLOPs 40.3704
MParams 177.9668
Source framework PyTorch*

Accuracy

Alphanumeric subset of common scene text recognition benchmarks are used. For your convenience you can see dataset size. Note, that we use here ICDAR15 alphanumeric subset without irregular (arbitrary oriented, perspective or curved) texts. See details here, section 4.1. All reported results are achieved without using any lexicon.

Dataset Accuracy Dataset size
ICDAR-03 92.96% 867
ICDAR-13 90.44% 1015
ICDAR-15 77.58% 1811
SVT 88.56% 647
IIIT5K 88.83% 3000

Input

Image, name: input, shape: 1, 1, 32, 100 in the format B, C, H, W, where:

  • B - batch size
  • C - number of channels
  • H - image height
  • W - image width

Note that the source image should be tight aligned crop with detected text converted to grayscale. Mean values: [127.5, 127.5, 127.5], scale factor for each channel: 127.5.

Outputs

Output tensor, name: output, shape: 1, 26, 37 in the format B, W, L, where:

  • W - output sequence length
  • B - batch size
  • L - confidence distribution across alphanumeric symbols: [s]0123456789abcdefghijklmnopqrstuvwxyz, where [s] - special end of sequence character for decoder.

The network output decoding process is pretty easy: get the argmax on L dimension, transform indices to letters and slice the resulting phrase on the first entry of end-of-sequence symbol.

Use text-detection demo

Model is supported by text-detection c++ demo(<omz_dir>/demos/text_detection_demo/cpp/main.cpp). In order to use this model in the demo, user should pass the following options:

  -tr_pt_first
  -dt "simple"

For more information, please, see documentation of the demo.

Demo usage

The model can be used in the following demos provided by the Open Model Zoo to show its capabilities:

Legal Information

The original model is distributed under the Apache License, Version 2.0. A copy of the license is provided in <omz_dir>/models/public/licenses/APACHE-2.0.txt.

[*] Other names and brands may be claimed as the property of others.