Doodle.AI: Redefining creative play with human-AI competition in real-time doodle recognition

Doodle.AI is an innovative web-based game that takes creativity, guessing, and artificial intelligence and puts them together in a great combination. It is both an entertaining and educational game, enabling players to enhance their creativity and guessing skills while exploring AI capabilities interactively. The participants draw random words as they compete with an artificial intelligence model, which predicts drawings in real time. The AI component not only adds challenge and thrill but also demonstrates the capability of machine learning in interactive gaming. To create a highly precise doodle recognizer, a Long-term Recurrent Convolutional Network (LRCN) was employed, and it proved to be much better than the typical Convolutional Neural Networks (CNNs). The LRCN model achieved 90 % accuracy in the top-1 score and 97 % in the top-3 score, which is a measure of its high level of prediction. Relative to CNNs, which were overfitting (training loss: 36 %, validation loss: 53 %), LRCN demonstrated a substantial learning curve, stabilizing at a final loss of 39 %, thereby offering better resilience. Furthermore, LRCN outperformed CNN, and therefore particularly has been found useful in dynamic gaming conditions. Doodle.AI is more than just for fun, providing interesting commentary on the domain of AI-based sketch recognition and the viability of deep learning algorithms in real time. The wide variety of categories and questions in the game makes it an enjoyable experience, as users are challenged to sharpen their creativity and intellectual skills while engaging with advanced AI technologies.

Authors

Institutions

Publication Details

Journal
Systems and Soft Computing
Published
2025-04-27
DOI
https://doi.org/10.1016/j.sasc.2025.200261
Primary Topic
Human Pose and Action Recognition
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Doodle.AI: Redefining creative play with human-AI competition in real-time doodle recognition

Biplov Paneru, Bishwash Paneru, Ishan Panta, Bipul Thapa
Systems and Soft Computing
Human Pose and Action Recognition
article

Doodle.AI: Redefining creative play with human-AI competition in real-time doodle recognition

Biplov Paneru, Bishwash Paneru, Ishan Panta, Bipul Thapa
article en

Abstract

Doodle.AI is an innovative web-based game that takes creativity, guessing, and artificial intelligence and puts them together in a great combination. It is both an entertaining and educational game, enabling players to enhance their creativity and guessing skills while exploring AI capabilities interactively. The participants draw random words as they compete with an artificial intelligence model, which predicts drawings in real time. The AI component not only adds challenge and thrill but also demonstrates the capability of machine learning in interactive gaming. To create a highly precise doodle recognizer, a Long-term Recurrent Convolutional Network (LRCN) was employed, and it proved to be much better than the typical Convolutional Neural Networks (CNNs). The LRCN model achieved 90 % accuracy in the top-1 score and 97 % in the top-3 score, which is a measure of its high level of prediction. Relative to CNNs, which were overfitting (training loss: 36 %, validation loss: 53 %), LRCN demonstrated a substantial learning curve, stabilizing at a final loss of 39 %, thereby offering better resilience. Furthermore, LRCN outperformed CNN, and therefore particularly has been found useful in dynamic gaming conditions. Doodle.AI is more than just for fun, providing interesting commentary on the domain of AI-based sketch recognition and the viability of deep learning algorithms in real time. The wide variety of categories and questions in the game makes it an enjoyable experience, as users are challenged to sharpen their creativity and intellectual skills while engaging with advanced AI technologies.

Systems and Soft ComputingVol. 7
Pokhara University (NP), Tribhuvan University (NP), Institute of Engineering (NP), Kathmandu University (NP)
Openalex Percentile: Top 93%
Human Pose and Action Recognition
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Doodle.AI: Redefining creative play with human-AI competition in real-time doodle recognition — Biplov Paneru, Bishwash Paneru, et al. · Systems and Soft Computing (2025) | TGRS Research Map | TGRS