I&T Solution |
Analyzing Health Conditions of Critical Lift Components with the Use of AI-Assisted Image Analytics
(REF: S-1363) |
Trial Project |
|
Solution Feature |
- Applying data augmentation technique to learn different variations of the same image, which can improve its performance.
- Use active learning to reduce the burden of manual data labeling by selecting the most informative images for labeling.
- Use Synthetic Data create additional training images that are not part of the original dataset when dealing with limited training data.
- Applying Pseudo Labeling (Semi-supervised learning) to help improve the model's performance.
|
Trial Application and Expected Outcome |
- Trained AI model that can accurately classify the site photos into 12 specific components or exceptions
- Trained AI model that can provide defect detection for 2 components (Suspension Rope and Brake)
- Develop and fintune model with overall of 90% or above accuracy
- Develop a scalable and adaptable AI model that can be applied new data for training and analysing via API
|
Info on I&T Solution Provider |
Solution Provider | : | SOCIF LIMITED | Address | : | Unit 527, 5/F, 5W Enterprise Place, Phase 1 Hong Kong Science Park, Pak Shek Kok, NT, HK | Contact Person | : | Jason Yuen |
Position | : | Chief Executive Officer | Tel | : | 96649906 | Email | : |
jason.yuen@socif.co |
|