Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
August 13, 2026Journal of Intelligent and Connected VehiclesOpen Access

Transformer-based out-of-distribution (OOD) input detection for trustworthy self-parking

View Full Paper
Ask AI
Bookmark
Share

Discussion

Loading...

Member takes

Implication

Randomized trial demonstrates enhanced OOD detection in self-parking systems, suggesting improved safety and efficiency.

Key Points

  • This research aims to improve the detection of out-of-distribution inputs in automated driving systems to enhance safety and reliability.
  • Enhanced a deep-reinforcement-learning-based self-parking agent with a novel OOD detector called TFMe.
  • Calibrated decision thresholds based on in-distribution data using a 99th-percentile rule.
  • Assessed sensitivity across various noise and attack types in real-world scenarios.
  • TFMe-enhanced agent mitigates collision risk and reduces timeouts in parking scenarios through timely Take-Over Requests.
  • Demonstrated 19% lower inference time and 30% lower energy consumption when deployed on Jetson Orin Nano compared to Raspberry Pi 5.
  • Achieved consistent performance across unseen anomaly types and parking environments without retraining.

Cite This Study

A 2026 study studied this question.

synapsesocial.com/papers/6a7d75d82b0e0cff3f63ec85https://doi.org/10.26599/jicv.2026.9210093
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Optimal deep transfer learning enabled object detector for anomaly recognition in pedestrian ways2024 · 2 citations
  2. 2Unseen Hazard Recognition in Autonomous Driving Using Vision–Language and Sensor-Based Temporal Models2026 · 1 citations
  3. 3Mitigating Overconfidence in Out-of-Distribution Detection by Capturing Extreme Activations2024
  4. 4Building Real-time Awareness of Out-of-distribution in Trajectory Prediction for Autonomous Vehicles2026 · 2 citations
  5. 5Language-Enhanced Latent Representations for Out-of-Distribution Detection in Autonomous Driving2024