MemoryPilotNet
These models contain end-to-end imitation learning-based models for control of an autonomous ego vehicle in CARLA simulator.
- GPU: NVIDIA GeForce 3090
- CUDA: 11.6
- Driver version: 510.54
- Input shape (150,50,3)
Tensorflow
- Tensorflow version: 2.7.0 / 2.11.0
| Model name | Visual memory | Kinematic memory | MAE test | MSE test | Model size (MB) | Filename | Previous filename |
|---|---|---|---|---|---|---|---|
| PilotNet | ✗ | ✗ | 0.048 | 0.018 | 122 | pilotnet_no_visual_no_kinematic.h5 | 20230125-120238_pilotnet_CARLA_14_12_dataset_bird_eye_300_epochs_no_flip_3_output_both_directions_all_towns_adam_AFFINE_PAPER.h5 |
| DeepestLSTMTinyPilotNet | ✗ | ✗ | 0.076 | 0.023 | 0.9 | deepestLSTMtinypilotnet_no_visual_no_kinematic.h5 | 20230126-183457_deepestLSTMTinyPilotNet_CARLA_14_12_dataset_bird_eye_400_epochs_no_flip_3_output_both_directions_all_towns__less_learning_rate_PAPER_cp.h5 |
| PilotNetx3 (TimeDistributed) | ✓ | ✗ | 0.028 | 0.0074 | 15.1 | pilotnet_time_distributed_visual_no_kinematic.h5 | 20230220-105422_pilotnet_small_more_more_LSTM_x3_CARLA_14_12_dataset_bird_eye_300_epochs_no_flip_3_output_both_directions_all_towns_sequences_more_more_extreme_t_5_t_10_dataset_AFFINE_PAPER_cp.h5 |
| PilotNetx3 (Conv3D) | ✓ | ✗ | 0.026 | 0.0091 | 127 | pilotnet_conv3d_visual_no_kinematic.h5 | 20230222-110851_memDCCP_small_LSTM_2_CARLA_14_12_dataset_bird_eye_300_epochs_no_flip_3_output_both_directions_all_towns_sequences_more_more_extreme_AFFINE_PAPER_cp.h5 |
| PilotNet | ✗ | ✓ | 0.031 | 0.0093 | 122 | pilotnet_no_visual_kinematic.h5 | 20230127-180655_pilotnet_CARLA_17_10_dataset_bird_eye_300_epochs_no_flip_3_output_velocity_all_towns_vel_30_AFFINE.h5 |
| DeepestLSTMTinyPilotNet | ✗ | ✓ | 0.065 | 0.016 | 0.9 | deepestLSTMtinypilotnet_no_visual_kinematic.h5 | 20230127-180856_deepestLSTMTinyPilotNet_CARLA_17_10_dataset_bird_eye_300_epochs_no_flip_3_output_velocity_all_towns_vel_30_AFFINE_PAPER.h5 |
| PilotNetx3 (TimeDistributed) | ✓ | ✓ | 0.0065 | 0.0083 | 15.1 | pilotnet_time_distributed_visual_kinematic.h5 | 20230221-162112_pilotnet_x3_small_CARLA_14_12_dataset_bird_eye_300_epochs_no_flip_3_output_velocity_both_directions_all_towns_sequences_more_extreme_previous_V_AFFINE_PAPER_cp.h5 |
| PilotNetx3 (Conv3D) | ✓ | ✓ | 0.0065 | 0.00077 | 127 | pilotnet_conv3d_visual_kinematic.h5 | 20230221-175351_memDCCP_small_CARLA_14_12_dataset_bird_eye_300_epochs_no_flip_3_output_velocity_both_directions_all_towns_sequences_more_extreme_previous_V_AFFINE_PAPER_cp.h5 |