Video Authoring Tool for Simulating Expert Perceptual Training
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current training tools for developing perceptual skills in experts and novices lack the ability to simulate expert experiences effectively without the presence of an expert, failing to provide adequate feedback on unprompted selections and rationales in real-time scenarios.
Innovation Solution
A computer-based authoring and simulation tool that allows experts to create processed videos with tagged objects and associated rationales, enabling users to make unprompted selections and enter freeform rationales, with the system providing expert rationales for comparison and accuracy scoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If current training tools are used to train novices on perceptual skills, then training can be provided, but the tools fail to simulate expert experiences effectively and provide inadequate feedback on unprompted selections
Solution Approach 1:
The patent creates a digital copy of expert decision-making processes by capturing expert selections and rationales during video playback, then using these copies to provide feedback to trainees. The system records expert annotations, selections, and explanations, storing them as reference data that automatically compares with and feedbacks to user selections, effectively simulating expert presence without requiring the actual expert to be present during training sessions.
2Measurement precision
If experts are present during training to provide feedback on perceptual skills, then accurate feedback is provided, but the training cannot be scaled and is time-consuming
Solution Approach 1:
The system enables self-service training by pre-capturing expert knowledge during video annotation phases, then allowing trainees to independently interact with the training material and receive automated feedback based on their selections. The expert has already performed the knowledge-capture work during video creation, and the system automatically serves this knowledge to multiple trainees simultaneously without requiring expert presence during each training session, thus achieving both accuracy and scalability.
Solution Approach 2:
The patent performs preliminary action by having experts annotate videos and capture their decision-making processes before the actual training sessions. All expert selections, annotations, and rationales are captured in advance during video creation, preparing the feedback mechanism beforehand. This preliminary capture of expert knowledge allows the system to provide accurate feedback during training without requiring real-time expert involvement, enabling scalable training delivery.
3Ease of operation
If multiple-choice questions are used to train novices, then structured feedback is provided, but the training does not capture unprompted selections or freeform rationales from users
Solution Approach 1:
The patent implements a dynamic feedback system that adapts to user interactions. When users make selections during video playback, the system dynamically determines whether to prompt for additional information based on selection accuracy. If selections are incorrect or incomplete, the system dynamically prompts users to enter freeform rationales, transforming from a static multiple-choice format to an interactive dialogue that captures both structured selections and unstructured reasoning, thereby preserving comprehensive user response data.
Data Source
AI summary
A process for simulating an expert experience comprises playing an expert simulation for a user, where the expert simulation is a processed video with tagged objects and untagged objects. During playback, a user may make an unprompted selection of an object in the video of the expert simulation, and the unprompted selection is received. In response to receiving the unprompted selection from the user, an entry interface for the user to enter in a freeform rationale is provided. After the user enters a freeform rationale through the entry interface, the freeform rationale is received. An expert rationale associated with a tagged object is displayed. Further, a process for creating the expert simulation comprises receiving raw video and superimposing a grid onto the raw video. An author tags objects within the video spatially, temporally, or both. Further, freeform data (e.g., an expert rationale) is associated with the tagged object.