Table 1 ROC scores in different datasets
From: A neural computational model for bottom-up attention with invariant and overcomplete representation
| Â | Â | Basic number | ROC | Â | Â | Basic number | ROC |
|---|---|---|---|---|---|---|---|
| Â | Â | (feature maps) | Â | Â | Â | (feature maps) | Â |
| Â | Â | 100(16) | 0.5722 | Â | Â | 196(25) | 0.7104 |
| Â | our | 196(25) | 0.6427 | Â | our | Â | Â |
Bruce | model | 392(50) | 0.6830 | Weizmann | model | 256(36) | 0.7282 |
data | Â | 576(64) | 0.6864 | data | Â | Â | Â |
| Â | Sal Tool | 144(30) | 0.6717 | Â | Sal Tool | 144(30) | 0.6233 |
| Â | ful net | 392(1) | 0.5913 | Â | ful net | 256(1) | 0.4947 |
| Â | rand net | 392(50) | 0.5965 | Â | rand net | 256(36) | 0.5011 |