Bayesian Metric Learning For Uncertainty Quantification In Image Retrieval

Abstract

We propose the first Bayesian encoder for metric learning. Rather than relying on neural amortization as done in prior works, we learn a distribution over the network weights with the Laplace Approximation. We actualize this by first proving that the contrastive loss is a valid log-posterior. We then propose three methods that ensure a positive definite Hessian. Lastly, we present a novel decomposition of the Generalized Gauss-Newton approximation. Empirically, we show that our Laplacian Metric Learner (LAM) estimates well-calibrated uncertainties, reliably detects out-of-distribution examples, and yields state-of-the-art predictive performance.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keywarburg2023bayesian

Related papers

• Bayesian Triplet Loss: Uncertainty Quantification In Image Retrieval (2020)11.49
• Hyp-uml: Hyperbolic Image Retrieval With Uncertainty-aware Metric Learning (2023)0.00
• Metric Learning With HORDE: High-order Regularizer For Deep Embeddings (2019)12.81
• Unsupervised Data Uncertainty Learning In Visual Retrieval Systems (2019)0.00
• Variance & Greediness: A Comparative Study Of Metric-learning Losses (2026)0.00
• Calibrated Neighborhood Aware Confidence Measure For Deep Metric Learning (2020)0.00
• Exploring Uncertainty Measures For Image-caption Embedding-and-retrieval Task (2019)2.26
• Large-scale Distance Metric Learning With Uncertainty (2018)9.03