Data Сollection for Сity Administration
Data collection improved the model’s performance in various street conditions by 85%.
- Industry and use case:
- Public Utilities
- Data:
- 7000 images
Challenge:
The city administration needed a comprehensive dataset featuring waste bins of various sizes and types, captured under different weather conditions and at various times of the day. The goal was to train a neural network to monitor timely waste collection and optimize vehicle logistics for waste disposal.
Solution:
Recognizing the importance of diverse conditions in the dataset, we opted to gather 7,000 images through a crowdsourcing initiative. This approach ensured a wide variety of scenarios, from day to night and across different weather conditions.
To meet the specific requirements, we also assembled a rapid-response team. This team followed designated routes through residential areas, capturing the necessary content regardless of challenging weather, including rain or nighttime conditions.
Results:
The client’s model saw an 85% improvement in performance across various street conditions.
This initiative enhanced the responsiveness of public services in addressing full waste bins.
Our Cases
-
Within a month, more than 10% of the entire database with over 2,000 photographs was collected
-
Through data collection, the client improved their biometric system for facial and voice recognition by 21%.
-
Content moderation on the video hosting platform enabled a 99% reduction in the influx of prohibited content
