Title | ||
---|---|---|
Two-Step Fine-Tuned Convolutional Neural Networks for Multi-label Classification of Children's Drawings |
Abstract | ||
---|---|---|
Developmental psychologists employ several drawing-based tasks to measure the cognitive maturity of a child. Manual scoring of such tests is time-consuming and prone to scorer bias. A computerized analysis of digitized samples can provide efficiency and standardization. However, the inherent variability of hand-drawn traces and lack of sufficient training samples make it challenging for both feature engineering and feature learning. In this paper, we present a two-step fine-tuning based method to train a multi-label Convolutional Neural Network (CNN) architecture, for the scoring of a popular drawing-based test 'Draw-A-Person' (DAP). Our proposed two-step fine-tuned CNN architecture outperforms conventional pre-trained CNNs by achieving an accuracy of 81.1% in scoring of Gross Details, 99.2% in scoring of Attachments, and 79.3% in scoring of Head Details categories of DAP samples. |
Year | DOI | Venue |
---|---|---|
2021 | 10.1007/978-3-030-86331-9_21 | DOCUMENT ANALYSIS AND RECOGNITION - ICDAR 2021, PT II |
Keywords | DocType | Volume |
Draw-a-person test, Multi-label classification, Two-step fine-tuning, Small sample domains | Conference | 12822 |
ISSN | Citations | PageRank |
0302-9743 | 0 | 0.34 |
References | Authors | |
0 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Muhammad Osama Zeeshan | 1 | 0 | 0.34 |
Imran Siddiqi | 2 | 421 | 36.56 |
Momina Moetesum | 3 | 1 | 1.70 |