SidewalkPilot-v3.4b
SidewalkPilot-v3.4b maps one 320x180 OpenCV BGR frame to steering control for the SidewalkPilot RC car. It is the best-validation checkpoint from the softened tree-shadow and left/right-balanced v3.4 run. It performed slightly worse than the final-epoch v3.4 checkpoint in physical testing and was not selected as the default.
Model Details
- Developer: Ram Shreyas Naik Sabavat
- Series: 3.x
- Architecture: SidewalkPilotV3 CNN, approximately 5.53 million parameters
- Model type: 9-class steering classification plus within-bucket offsets and throttle
- Checkpoint role: best-validation
- Field status: tested July 13, 2026; runner-up to v3.4
- Artifact:
SidewalkPilot-v3.4b.onnx(FP32 ONNX) - Checkpoint created: 2026-07-12 21:18 PDT
- Input:
[batch, 3, 180, 320], OpenCV BGR - Preprocessing: resize to
320x180, then(x / 255 - 0.5) / 0.5 - Output:
[batch, 19]: 9 class logits, 9 offsets, and throttle
Output Decoding
class_id = argmax(output[0:9])
fraction = sigmoid(output[9 + class_id])
steering_deg = bucket_low[class_id] + fraction * bucket_width[class_id]
throttle = sigmoid(output[18]) # not used by the current runtime
Logical steering uses 0 for full left, 90 for straight, and 180 for full right. The throttle output is not deployed because the collected throttle labels do not contain enough variation to train useful throttle control.
Evaluation Setup
- Evaluation set: 81,237 current Series 3 real field images
- Purpose: training-set fit check, not held-out generalization proof
- Error unit: logical steering degrees
- Selection priority: balanced 9-bucket accuracy and turn recall first; MAE is supporting evidence
- Cross-series warning: do not compare these numbers directly with Series 1/2, which use their original 2,224-image dataset
Chronological Metrics Through This Version
| Model | Bal9 | Turn exact | Turn +/-1 | Straight exact | MAE | Median AE | Signed error |
|---|---|---|---|---|---|---|---|
3.0 | 16.0% | 15.3% | 42.9% | 23.4% | 18.971 | 13.623 | -3.594 |
3.0b | 15.9% | 14.8% | 42.3% | 24.6% | 18.450 | 13.129 | -3.944 |
3.1 | 28.1% | 26.8% | 53.3% | 55.2% | 22.647 | 9.729 | -2.353 |
3.1b | 27.4% | 25.8% | 52.3% | 56.7% | 20.958 | 9.591 | -3.769 |
3.2 | 34.0% | 31.1% | 57.7% | 52.4% | 16.776 | 9.648 | -1.474 |
3.2b | 25.1% | 19.8% | 46.0% | 67.2% | 14.457 | 5.909 | -4.691 |
3.3 | 27.1% | 21.2% | 48.1% | 68.5% | 14.630 | 5.204 | -5.351 |
3.3b | 19.2% | 10.6% | 36.5% | 78.0% | 16.460 | 2.183 | -8.287 |
3.4 | 33.3% | 32.6% | 60.7% | 65.5% | 14.377 | 4.859 | +0.799 |
3.4b (this model) | 25.6% | 22.7% | 52.2% | 73.7% | 13.904 | 2.678 | -2.093 |
Class-Balanced Ranking Through This Version
| Rank | Model | Bal9 | Turn exact | Turn +/-1 | MAE | Median AE | Signed error |
|---|---|---|---|---|---|---|---|
1 | 3.2 | 34.0% | 31.1% | 57.7% | 16.776 | 9.648 | -1.474 |
2 | 3.4 | 33.3% | 32.6% | 60.7% | 14.377 | 4.859 | +0.799 |
3 | 3.1 | 28.1% | 26.8% | 53.3% | 22.647 | 9.729 | -2.353 |
4 | 3.1b | 27.4% | 25.8% | 52.3% | 20.958 | 9.591 | -3.769 |
5 | 3.3 | 27.1% | 21.2% | 48.1% | 14.630 | 5.204 | -5.351 |
6 | 3.4b | 25.6% | 22.7% | 52.2% | 13.904 | 2.678 | -2.093 |
7 | 3.2b | 25.1% | 19.8% | 46.0% | 14.457 | 5.909 | -4.691 |
8 | 3.3b | 19.2% | 10.6% | 36.5% | 16.460 | 2.183 | -8.287 |
9 | 3.0 | 16.0% | 15.3% | 42.9% | 18.971 | 13.623 | -3.594 |
10 | 3.0b | 15.9% | 14.8% | 42.3% | 18.450 | 13.129 | -3.944 |
Current Version Snapshot
- Balanced 9-bucket exact: 25.6%
- Turn exact: 22.7%
- Turn within one bucket: 52.2%
- Straight exact: 73.7%
- MAE: 13.904 degrees
- Median absolute error: 2.678 degrees
- Signed error: -2.093 degrees
Field Evaluation
v3.4b was tested on the physical car on July 13, 2026 against v3.4, v3.3, and v3.3b. The comparison included harsh-shadow cases and normal left/right turns. The operator reported that v3.4b was slightly worse than the regular v3.4 checkpoint, so v3.4 was selected for the runtime.
v3.4b's lower MAE did not outweigh v3.4's stronger balanced and turn metrics or the physical result. The field verdict is qualitative because the exact route, weather, duration, manual-takeover count, and linked clip identifiers were not preserved.
Intended Use
- Small RC-car autonomy experiments
- Sidewalk/path steering research
- Jetson ONNX/TensorRT deployment experiments
Limitations and Safety
This model does not identify obstacles, prove a clear path, estimate confidence, or detect out-of-distribution scenes. Shadows, lighting changes, curb geometry, driveways, camera movement, and underrepresented turns can cause unsafe steering. Keep independent LiDAR emergency braking, manual takeover, conservative speed limits, and bounded operating conditions above model output. Its July 13 runner-up result is limited to that comparison and is not a general safety claim.
Reproducibility
artifact_manifest.json records the artifact SHA-256, tensor signature, checkpoint role, evaluation set, and current report metrics.