Algorithm 1.
Intrinsic phase: one step of learning of affordances and forward models
| 1: | (object_image, object_position) ← Scan(environment) | |
| 2: | (action, motivation_signal) ← SelectActionWithHighestIM(action_list, predictors, object_image, object_position) | |
| 3: | if (motivation_signal ≥ motivation_threshold) then | |
| 4: | ExecuteAction(action, object_image, object_position) | |
| 5: | (new_object_image, new_object_position) ← ScanEffect(new_environment, environment) | |
| 6: | affordance ←… | |
| 7: | Affordance(action, new_object_image, new_object_position, object_image, object_position) | |
| 8: | UpdateWeights(affordance_predictor, action, object_image, affordance) | |
| 9: | UpdateWeights(affordance_predictor, action, object_image, affordance, improve_predictor) | ⊳ Only IMP |
| 10: | if (affordance = TRUE) then | |
| 11: | UpdateWeights(effect_predictors, action, object_image, object_position,… | |
| 12: | new_object_image, new_object_position) | |
| 13: | end if | |
| 14: | motivation_threshold ← LeakyAverage(motivation_threshold, motivation_signal) | ⊳ Only IGN/IMP |
| 15: | else | |
| 16: | motivation_threshold ← LeakyAverage(motivation_threshold, 0) | ⊳ Only IGN/IMP |
| 17: | end if |