Method for freehand sketch training
Inventors
Assignees
National Science Foundation NSF • University of California San Diego UCSD
Publication Number
US-11645939-B2
Publication Date
2023-05-09
Expiration Date
2039-12-02
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Abstract
A computer-based method and system for teaching educational sketching includes receiving a user-generated image created in response to a learning assignment and comparing the user-generated image to a solution image to identify one or more errors in the user-generated image relative to the solution image, where the errors may include additional image elements and missing image elements. Comparing is performed by providing a solution region corresponding to an acceptable variation from the solution image and identifying one or more errors based on a presence or absence of at least a portion of a corresponding element of the user-generated image within the solution region. If errors are identified and a non-passing status is determined, a hint is displayed to the user. The hint may be the correct elements of the user-generated image, a portion of the solution image, or a combination thereof.
Core Innovation
The invention is a computer-based method and system for teaching educational sketching by receiving a user-generated image created in response to a learning assignment and comparing it to a solution image to identify errors such as additional or missing image elements. The comparison involves providing a solution region that corresponds to an acceptable variation from the solution image, and identifying errors based on whether portions of the user's image elements are within this solution region. If the errors exceed a threshold, a non-passing status is indicated, and a hint is generated and displayed to the user. This hint may include parts of the error, correct user elements, or parts of the solution image. The user is then prompted to submit a new image, and the process repeats until the user's image passes the threshold and is considered correct.
The method provides real-time feedback and grading in a manner similar to an expert teacher who accounts for degrees of correctness rather than binary pass/fail judgments. It distinguishes between intentional and unintentional marks to avoid being too lenient or too strict. Feedback includes options for retrying, receiving hints, or peeking at the solution, encouraging persistence and incremental learning rather than prematurely revealing answers. Hints and feedback images are visually encoded, often employing color coding to indicate correct, missing, or additional elements, thereby guiding the student toward the correct sketch.
The problem being solved addresses the limitations of traditional sketch teaching methods and computer-based multiple-choice assessments. Traditional pen and paper sketching requires delayed grading and limited instructor interaction, while multiple-choice methods fail to provide meaningful iterative attempts or customized feedback based on student work. There is a need for a technology-assisted method that provides immediate, customized feedback similar to an expert teacher, encourages student persistence, and supplies instructors with detailed information about student errors and effort. The invention responds to these problems by providing an automated grading system that accepts freehand sketches, assesses their correctness with tolerances, and provides targeted hints and feedback to enhance the learning experience and reduce instructor workload.
Claims Coverage
The claims define several inventive features embodied in a method and system for educational sketching that automatically compare user-generated images with a solution image to identify errors and provide feedback, including hints, evaluations, and repeated attempts until passing.
Automated error identification using solution regions
The method receives a user-generated image and compares it to a stored solution image by defining a solution region around each element of the solution that corresponds to acceptable variation. Errors are identified based on whether user-generated elements fall within or outside these solution regions, distinguishing between missing and additional image elements.
Evaluation of error magnitude against thresholds to determine pass status
After identifying errors, the method determines the magnitude of each error and compares these magnitudes to pre-set thresholds. A non-passing status is indicated if any error magnitude exceeds its threshold, enabling grading that accounts for degrees of correctness rather than binary pass/fail.
Provision of graphical hints tailored to user errors and progress
If a non-passing status is indicated, the system generates at least one graphical hint for display. Hints may comprise portions of the user's errors, correct user-generated elements, or parts of the solution image, with hints visually encoded to distinguish correct, missing, and additional elements, often using color coding to guide the user without revealing complete solutions.
Support for multiple input device interfaces
The system supports a variety of device interfaces including touchscreens, computer mice, trackballs, cameras, image scanners, and motion sensors to receive sketches, allowing flexible input modalities for user-generated images.
Specialized grading for solutions containing dashed lines
The method handles solution images with dashed lines by defining dashed line solution regions and identifies errors such as excessively long dashes or gaps beyond acceptable thresholds, thus accommodating technical drawings with dashed and solid line conventions.
Iterative evaluation with prompts for resubmission until passing
The method repeats the process of receiving user-generated images, grading, providing hints if needed, and prompting for resubmission until the sketch meets the passing threshold, thus fostering persistence and iterative learning.
The claims collectively describe an interactive educational sketching system and method that automatically grades user-generated sketches relative to solution images using defined solution regions, evaluates errors quantitatively against thresholds, provides tailored graphical hints and iterative feedback, supports diverse input devices, accommodates dashed line drawings, and encourages repeated user attempts until achieving correctness.
Stated Advantages
Provides immediate, customized feedback to students that emulates expert teacher grading by allowing partial correctness and distinguishing between intentional and unintentional marks.
Encourages student persistence by permitting multiple attempts and offering graduated hints rather than revealing complete solutions immediately.
Reduces instructor grading workload while enabling measurement of individual student effort and error types for targeted guidance.
Offers a more realistic, flexible alternative to multiple-choice question methods by supporting freehand sketch input and continuous feedback.
Improves clarity of feedback through color-coded and visually distinguishable indicators showing correct, missing, and additional sketch elements.
Documented Applications
Teaching spatial visualization and hand sketching skills in STEM fields such as engineering technical drawings including orthographic and isometric views.
Supporting educational sketching in broader contexts such as arithmetic teaching with block sets, math topics like ratios through pie charts, anatomy sketching for surgery students, and geology representations such as plate tectonics.
Providing feedback and grading for 2D and 3D sketches using various input devices like touchscreens, motion sensors, and scanners.
Facilitating persistence and iterative learning by enabling repeated sketch submissions with tailored hints and peeks at solution components.
Interested in licensing this patent?