Video Object Tracking for Automated Merchandise Recommendation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional methods for recommending video content-related merchandise during video playback are labor-intensive and prone to omitting similar merchandise due to the large volume of video content, leading to incomplete recommendations.

Innovation Solution

A system and process that utilize tracking recognition techniques to identify target objects in video frames, determine time slice information, and automatically match similar objects from a library, presenting information concurrently with video playback.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If manual labeling and matching of video content-related merchandise is used, then merchandise recommendations can be provided during video playback, but the process becomes extremely time-consuming and labor-intensive when video content is large in volume

Engineering Contradiction:

Improveease of merchandise recommendationVSAvoidtime consumption

Core Design Contradiction:

Ease of operationVSLoss of time

Solution Approach 1:

The patent replaces the manual mechanical labeling and matching process with an automated computer-based system that uses video analysis and object recognition technologies to automatically identify merchandise in video frames and match them with corresponding product information, thereby eliminating the time-consuming manual operations while maintaining recommendation accuracy

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables the video content itself to 'self-identify' merchandise through automated detection algorithms that analyze video frames, extract object features, and automatically generate merchandise recommendations without requiring external manual intervention, making the recommendation process self-sufficient and highly efficient

Inventive Principle:

Principle #25Self-service

2Reliability

If manual labeling and matching is used for a large number of videos, then some merchandise recommendations can be provided, but omissions of similar merchandise occur resulting in incomplete recommendations

Engineering Contradiction:

Improvecompleteness of merchandise recommendationVSAvoidoperational complexity

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The patent implements a dynamic recommendation system that continuously adapts to different video content by automatically adjusting its analysis parameters and matching criteria based on the specific characteristics of each video, allowing it to comprehensively identify various types of merchandise including similar items that static manual processes would miss

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system dynamically changes multiple parameters including object detection thresholds, feature extraction weights, and matching criteria based on the video content being analyzed, enabling it to accurately identify and recommend similar merchandise across diverse video types without manual reconfiguration

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11284168B2Presenting information on similar objects relative to a target object from a plurality of video frames

Publication Date: 2022.03.22 ALIBABA GROUP HOLDING LTD

AI summary

Presenting information on similar objects relative to a target object is disclosed, including: obtaining a plurality of video frames; determining a target object in the plurality of video frames using a tracking recognition technique; determining time slice information corresponding to the target object; using the time slice information corresponding to the target object to determine one or more similar objects relative to the target object; receiving an indication to present information on the one or more similar objects relative to the target object; and outputting the information on the one or more similar objects relative to the target object.