Training-Free Out-Of-Distribution Segmentation With Foundation Models
This addresses safety-critical applications like autonomous driving by enabling out-of-distribution segmentation with minimal assumptions, though it is incremental as it builds on existing foundation models.
The paper tackled detecting unknown objects in semantic segmentation without outlier supervision by proposing a training-free method using foundation model features and clustering, achieving 50.02 AP on RoadAnomaly and 48.77 AP on ADE-OoD benchmarks.
Detecting unknown objects in semantic segmentation is crucial for safety-critical applications such as autonomous driving. Large vision foundation models, including DINOv2, InternImage, and CLIP, have advanced visual representation learning by providing rich features that generalize well across diverse tasks. While their strength in closed-set semantic tasks is established, their capability to detect out-of-distribution (OoD) regions in semantic segmentation remains underexplored. In this work, we investigate whether foundation models fine-tuned on segmentation datasets can inherently distinguish in-distribution (ID) from OoD regions without any outlier supervision. We propose a simple, training-free approach that utilizes features from the InternImage backbone and applies K-Means clustering alongside confidence thresholding on raw decoder logits to identify OoD clusters. Our method achieves 50.02 Average Precision on the RoadAnomaly benchmark and 48.77 on the benchmark of ADE-OoD with InternImage-L, surpassing several supervised and unsupervised baselines. These results suggest a promising direction for generic OoD segmentation methods that require minimal assumptions or additional data.