In-Vehicle Conversation Topic Selection by Passenger Field of View
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing conversational systems in vehicles struggle to provide smooth conversations with onboard persons due to changing scenery and reduced concentration during travel, especially in autonomous driving, making it difficult for passengers to find relevant topics for discussion.
Innovation Solution
An information processing apparatus selects objects within a predetermined range visible to the onboard person from a captured image, generating utterance content based on these objects as topics, and adjusts the range based on vehicle speed and driving mode to ensure ease of recognition and reduce processing load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the conversational system selects objects from the entire captured image range, then the system can discuss any scenery, but the onboard person may not be able to easily find the topic and concentration is reduced
Solution Approach 1:
The patent divides the captured image into a predetermined range based on the onboard person's field of view, selecting objects only from this segmented region rather than the entire image. This ensures the conversation topics are within the person's visible area, making it easier for them to locate and engage with the discussion subject.
2Productivity
If the system processes the entire captured image, then all scenery can be analyzed, but the processing load increases
Solution Approach 1:
The patent extracts only the predetermined range from the captured image based on the onboard person's field of view, and processes objects solely within this extracted region. This reduces the processing load by focusing computational resources only on the relevant portion of the image rather than analyzing the entire captured scene.
3Ease of operation
If the system selects objects from a limited predetermined range, then the processing load is reduced and ease of recognition is improved, but the conversation topic coverage is limited
Solution Approach 1:
The patent dynamically adjusts the predetermined range based on the onboard person's field of view, which changes as the vehicle moves and the person's viewing angle changes. This dynamic adaptation ensures that the selected objects remain within the person's visible range while maximizing conversation topic coverage within the constraints of their field of view.
Data Source
AI summary
An information processing apparatus includes a processor configured to: select one or more first objects detected from a predetermined range in a captured image from a vehicle-mounted camera that takes an outward direction from a vehicle as a capturing range, the predetermined range being based on a field of view of an onboard person in the vehicle, and generate utterance content that takes the one or more first objects as a topic, the utterance content being to be uttered to the onboard person. The processor sets the predetermined range based on at least one of a traveling speed of the vehicle, whether traveling is performed by autonomous driving of the vehicle or not, and a position where a line of sight of the onboard person is focused on.