7th AI Driving Olympics: 1st Place Report for Panoptic Tracking

TL;DR

This paper presents EfficientLPT, a novel architecture that achieved first place in the panoptic tracking challenge at NeurIPS 2021, by effectively fusing multi-scale features and utilizing scan overlaps for consistent ID tracking.

Contribution

The paper introduces EfficientLPT, a new architecture with a panoptic fusion module and range-aware features, improving panoptic tracking performance in autonomous driving scenarios.

Findings

Achieved 1st place in the NeurIPS 2021 panoptic tracking challenge.

Outperformed existing methods on the Panoptic nuScenes dataset.

Demonstrated effective use of scan overlaps for consistent ID tracking.

Abstract

In this technical report, we describe our EfficientLPT architecture that won the panoptic tracking challenge in the 7th AI Driving Olympics at NeurIPS 2021. Our architecture builds upon the top-down EfficientLPS panoptic segmentation approach. EfficientLPT consists of a shared backbone with a modified EfficientNet-B5 model comprising the proximity convolution module as the encoder followed by the range-aware FPN to aggregate semantically rich range-aware multi-scale features. Subsequently, we employ two task-specific heads, the scale-invariant semantic head and hybrid task cascade with feedback from the semantic head as the instance head. Further, we employ a novel panoptic fusion module to adaptively fuse logits from each of the heads to yield the panoptic tracking output. Our approach exploits three consecutive accumulated scans to predict locally consistent panoptic tracking IDs and also the overlap between the scans to predict globally consistent panoptic tracking IDs for a given sequence. The benchmarking results from the 7th AI Driving Olympics at NeurIPS 2021 show that our model is ranked #1 for the panoptic tracking task on the Panoptic nuScenes dataset.