In-Place Scrolling Interface Using Neural Feature Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional user interfaces for online marketplaces limit users to viewing one item at a time, requiring scrolling to browse multiple items, which can be inefficient and disorienting, especially on devices with limited screen real estate, as users must stop scrolling to comprehend images and may not find similar items easily.
Innovation Solution
The in-place scrolling feature uses a deep neural network to generate feature vectors for images, allowing similar items to be transitioned in-place within the user interface without moving the focus location, ensuring a smooth and efficient browsing experience by selecting candidate images based on Euclidean distance and other factors like relevance and user history.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If conventional scrolling interface is used to browse multiple items, then users can view more items, but users must stop scrolling to comprehend images and lose focus location
Solution Approach 1:
The interface is segmented into a focus location (center) and peripheral locations. Only one item is displayed at the focus location at a time, while other items are displayed in peripheral locations. This segmentation allows users to comprehend the focused item fully while having access to other items without scrolling.
Solution Approach 2:
The patent transitions from one-dimensional scrolling (vertical or horizontal movement through a list) to a two-dimensional spatial arrangement where items are positioned in peripheral locations around the focus location. This dimensional change allows multiple items to be visible simultaneously without requiring scroll movement, resolving the contradiction between viewing quantity and comprehension ease.
2Adaptability or versatility
If conventional scrolling interface is used, then users can browse through items, but users must provide continuous scroll input which is inefficient on devices with limited screen real estate
Solution Approach 1:
The patent merges the functions of scrolling (navigation through multiple items) and focusing (comprehending individual items) into a single interface state. Users can view multiple items simultaneously in peripheral locations while maintaining focus on one item in the center, eliminating the need for continuous scroll input and improving browsing efficiency on devices with limited screen real estate.
Solution Approach 2:
The system prepares and displays multiple candidate items in peripheral locations in advance, so when users want to browse, the items are already visible and ready for selection without requiring scroll input to load or navigate to them. This preliminary arrangement of items resolves the contradiction between browsing capability and efficiency.
3Adaptability or versatility
If conventional scrolling interface is used, then users can navigate through items, but users cannot easily find similar items without scrolling through all items
Solution Approach 1:
The system uses machine learning models to analyze the focused item and provide feedback by identifying and displaying similar items in peripheral locations. This feedback mechanism allows users to quickly find similar items without scrolling through unrelated items, reducing the time lost in navigation while maintaining versatile item browsing capability.
Solution Approach 2:
The patent replaces the mechanical scrolling system (manual navigation through items) with an intelligent system using machine learning models that automatically identify and present similar items. This substitution eliminates the need for users to manually scroll through items to find similar ones, significantly reducing the time required while maintaining navigation capability.
Data Source
AI summary
Techniques for providing an in-place scrolling feature are described herein. Input via a user interface may be received to update first images of first items to second images of second items where the first images are presented in a focus location of the user interface. Feature vectors may be calculated for each image associated with the first items and the second items using a neural network algorithm. The feature vectors may include a plurality of values that represent each image of the first items and the second items. Candidate images may be determined from the second images based on a distance in a feature vector space between the feature vectors of the candidate images and the feature vectors of the second images. The user interface may be updated to transition from presenting the first images to a portion of the candidate images in the focus location of the user interface.