AI algorithms performance assessment
Summary
AI algorithms performance assessment
S00124
Version: 1.11
- horticulture
- tree crops
- viticulture
- performance evaluation
- software or AI model
- Location: at user's premises, in Italy, remote
- Offered by: FBK
Description
Support clients in defining metrics and testing their AI solutions in real life, in the field or virtually. Within normal and "stressed" operating conditions, including AI generated test cases, we assess the readiness of agriculture related AI algorithms for classification (weeds, chromatic analysis of crops), surround understanding and perception capabilities (for safety, for navigation), yield measurement (quantity of produce, quality), robotics (anomaly detection, etc.). Besides accuracy metrics, we have experience evaluating Memory and computing power footprint , and by doing so understanding bottlenecks and improving energy efficiency of proposed solutions.