What is wrong with scene text recognition model comparisons? dataset and model analysis

04/03/2019
by   Jeonghun Baek, et al.
0

Many new proposals for scene text recognition (STR) models have been introduced in recent years. While each claim to have pushed the boundary of the technology, a holistic and fair comparison has been largely missing in the field due to the inconsistent choices of training and evaluation datasets. This paper addresses this difficulty with three major contributions. First, we examine the inconsistencies of training and evaluation datasets, and the performance gap results from inconsistencies. Second, we introduce a unified four-stage STR framework that most existing STR models fit into. Using this framework allows for the extensive evaluation of previously proposed STR modules and the discovery of previously unexplored module combinations. Third, we analyze the module-wise contributions to performance in terms of accuracy, speed, and memory demand, under one consistent set of training and evaluation datasets. Such analyses clean up the hindrance on the current comparisons to understand the performance gain of the existing modules.

READ FULL TEXT

page 3

page 8

page 10

research
07/25/2021

Comprehensive Studies for Arbitrary-shape Scene Text Detection

Numerous scene text detection methods have been proposed in recent years...
research
05/18/2023

a unified front-end framework for english text-to-speech synthesis

The front-end is a critical component of English text-to-speech (TTS) sy...
research
05/17/2021

STRIDE : Scene Text Recognition In-Device

Optical Character Recognition (OCR) systems have been widely used in var...
research
11/04/2019

Scene Text Recognition with Temporal Convolutional Encoder

Texts from scene images typically consist of several characters and exhi...
research
08/30/2018

LUCSS: Language-based User-customized Colourization of Scene Sketches

We introduce LUCSS, a language-based system for interactive col- orizati...
research
12/19/2022

Statistical Dataset Evaluation: Reliability, Difficulty, and Validity

Datasets serve as crucial training resources and model performance track...
research
11/19/2014

Ontology Module Extraction via Datalog Reasoning

Module extraction - the task of computing a (preferably small) fragment ...

Please sign up or login with your details

Forgot password? Click here to reset