Pedestrian crossing intention prediction accuracy for different methods on three public datasets.
| Method | PIE | JAAD all | JAAD beh | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ACC | AUC | F1 | Predskιn | Recall | ACC | AUC | F1 | Precision | Recall | ACC | AUC | F1 | Precision | Recall | |
| SFGRU ([23]) | 0.82 | 0.79 | 0.69 | 0.67 | 0.7 | 0.84 | 0.84 | 0.65 | 0.54 | 0.84 | 0.51 | 0.45 | 0.63 | 0.61 | 0.64 |
| I3D ([7]) | 0.81 | 0.83 | 0.72 | 0.60 | 0.9 | 0.84 | 0.8 | 0.63 | 0.55 | 0.73 | 0.62 | 0.51 | 0.75 | 0.65 | 0.88 |
| TrouSPI-Net ([9]) | 0.88 | 0.87 | 0.80 | 0.77 | 0.84 | 0.82 | 0.77 | 0.58 | 0.49 | 0.70 | 0.64 | 0.55 | 0.76 | 0.65 | 0.91 |
| MultiRNN ([4]) | 0.83 | 0.8 | 0.71 | 0.69 | 0.73 | 0.79 | 0.79 | 0.58 | 0.45 | 0.79 | 0.61 | 0.5 | 0.74 | 0.64 | 0.86 |
| D. Yang. et al. ([40]) | - | - | - | - | - | 0.83 | 0.82 | 0.63 | 0.51 | 0.81 | 0.62 | 0.54 | 0.74 | 0.65 | 0.85 |
| PCPA ([13]) | 0.87 | 0.86 | 0.77 | - | - | 0.85 | 0.86 | 0.68 | - | - | 0.58 | 0.5 | 0.71 | - | - |
| IntFormer ([16]) | 0.89 | 0.92 | 0.81 | - | - | 0.86 | 0.78 | 0.62 | - | - | 0.59 | 0.54 | 0.69 | - | - |
| Yu Yao. et al. ([41]) | 0.84 | 0.90 | 0.88 | 0.96 | - | 0.87 | 0.70 | 0.92 | 0.66 | - | - | - | - | - | - |
| BiPed ([25]) | 0.91 | 0.90 | 0.85 | 0.82 | - | 0.83 | 0.79 | 0.60 | 0.52 | - | - | - | - | - | - |
| Ours | 0.91 | 0.89 | 0.84 | 0.84 | 0.85 | 0.89 | 0.78 | 0.66 | 0.72 | 0.61 | 0.68 | 0.63 | 0.76 | 0.71 | 0.81 |
| Method | PIE | JAAD all | JAAD beh | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ACC | AUC | F1 | Predskιn | Recall | ACC | AUC | F1 | Precision | Recall | ACC | AUC | F1 | Precision | Recall | |
| SFGRU ([ | 0.82 | 0.79 | 0.69 | 0.67 | 0.7 | 0.84 | 0.84 | 0.65 | 0.54 | 0.51 | 0.45 | 0.63 | 0.61 | 0.64 | |
| I3D ([ | 0.81 | 0.83 | 0.72 | 0.60 | 0.84 | 0.8 | 0.63 | 0.55 | 0.73 | 0.62 | 0.51 | 0.75 | 0.65 | 0.88 | |
| TrouSPI-Net ([ | 0.88 | 0.87 | 0.80 | 0.77 | 0.84 | 0.82 | 0.77 | 0.58 | 0.49 | 0.70 | 0.64 | 0.55 | 0.76 | 0.65 | |
| MultiRNN ([ | 0.83 | 0.8 | 0.71 | 0.69 | 0.73 | 0.79 | 0.79 | 0.58 | 0.45 | 0.79 | 0.61 | 0.5 | 0.74 | 0.64 | 0.86 |
| D. Yang. et al. ([ | - | - | - | - | - | 0.83 | 0.82 | 0.63 | 0.51 | 0.81 | 0.62 | 0.54 | 0.74 | 0.65 | 0.85 |
| PCPA ([ | 0.87 | 0.86 | 0.77 | - | - | 0.85 | 0.68 | - | - | 0.58 | 0.5 | 0.71 | - | - | |
| IntFormer ([ | 0.89 | 0.81 | - | - | 0.86 | 0.78 | 0.62 | - | - | 0.59 | 0.54 | 0.69 | - | - | |
| Yu Yao. et al. ([ | 0.84 | 0.90 | - | 0.87 | 0.70 | 0.66 | - | - | - | - | - | - | |||
| BiPed ([ | 0.90 | 0.85 | 0.82 | - | 0.83 | 0.79 | 0.60 | 0.52 | - | - | - | - | - | - | |
| Ours | 0.89 | 0.84 | 0.84 | 0.85 | 0.78 | 0.66 | 0.61 | 0.81 | |||||||
Sharing content requires targeting cookies to be enabled. Please update your cookie preferences to use this feature.