SPK：誘發結構化先驗知識以實現即時物件偵測中的可解釋分布外偵測

摘要

物體偵測器經常對其訓練類別之外的物體產生過度自信的預測，從而導致所謂的分布外（OoD）幻覺。現有的偵測或緩解此類幻覺的方法，通常要么直接在學習到的物體偵測器表徵上建構評分函數，要么修改物體偵測器本身以抑制幻覺的產生。然而，這些表徵中隱式編碼的潛在先驗在很大程度上仍未被探索，也尚未被明確解碼用於 OoD 偵測。為了揭示並利用這些潛在先驗，我們提出了結構化先驗知識（SPK），這是一個以幻覺為導向的框架，明確地從預訓練的物體偵測器中引出與 OoD 相關的先驗。具體而言，SPK 利用分布內數據和誘發幻覺的樣本作為診斷性監督，以引出支撐物體偵測器決策的部件級語義概念，而非僅將其用於拒絕或物體偵測器的適應。所引出的語義先驗進一步與幾何先驗和上下文先驗整合，形成緊湊的五維 SPK 表徵，用於 OoD 偵測。跨越多種物體偵測器架構和多個 OoD 基準的大量實驗表明，SPK 達到了最先進的 OoD 偵測性能。我們的研究結果揭示，預訓練的物體偵測器所編碼的潛在知識，遠比通常被利用於 OoD 偵測的知識要豐富得多。更重要的是，這些知識可以被明確引出並組織成一個緊湊、結構化且可解釋的知識空間，用於預測可靠性分析。這為通過明確揭示和利用潛在先驗來提高物體偵測器的可靠性，指明了一條有前景的主動式途徑。代碼和數據可在以下網址獲取：https://gricad-gitlab.univ-grenoble-alpes.fr/dnn-safety/spk

English

Object detectors often produce over-confident predictions for objects outside their training categories, leading to so-called out-of-distribution (OoD) hallucinations. Existing approaches for detecting or mitigating such hallucinations typically either construct scoring functions directly over learned object detector representations or modify the object detector itself to suppress hallucination emergence. However, the latent priors implicitly encoded in these representations remain largely unexplored and have not been explicitly decoded for OoD detection. To uncover and exploit these latent priors, we propose Structured Prior Knowledge (SPK), a hallucination-oriented framework that explicitly elicits OoD-relevant priors from pretrained object detectors. Specifically, SPK leverages in-distribution data and hallucination-inducing samples as diagnostic supervision to elicit part-level semantic concepts underlying object detector decision-making, rather than using them merely for rejection or object detector adaptation. The elicited semantic priors are further integrated with geometric and contextual priors to form a compact five-dimensional SPK representation for OoD detection. Extensive experiments across diverse object detector architectures and multiple OoD benchmarks demonstrate that SPK achieves state-of-the-art OoD detection. Our findings reveal that pretrained object detectors already encode substantially richer latent knowledge than is typically exploited for OoD detection. More importantly, this knowledge can be explicitly elicited and organized into a compact, structured, and interpretable knowledge space for prediction reliability analysis. This suggests a promising proactive route for improving object detector reliability by explicitly uncovering and leveraging latent priors. Code and data are available at: https://gricad-gitlab.univ-grenoble-alpes.fr/dnn-safety/spk