Rethinking FID Through the Geometry of the Reference Dataset
Fréchet Inception Distance (FID) is widely used to evaluate image generators, yet lower FID does not always correspond to better sample quality. We show that this mismatch depends in part on the geometry of the reference dataset. In a controlled study across six datasets, distributional density and effective rank significantly explain how FID changes as sample quality improves. Concentrated datasets tend to yield more favorable FID trends, whereas more dispersed datasets can make FID worsen despite better samples. Attribution to precision and recall and ablations with alternative feature spaces and distances support the same conclusion. These results suggest that distributional metrics should be interpreted together with the geometry of the reference dataset for more reliable benchmarking.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Rethinking Popularity Bias in Collaborative Filtering via Analytical Vector Decomposition
Popularity bias fundamentally undermines the personalization capabilities of collaborative filtering (CF) models, causing them to disproportionately recommend popular items while neglecting users' genuine preferences for…
Collaborative FilteringRethinking LLM Training through Information Geometry and Quantum Metrics
Optimization in large language models (LLMs) unfolds over high-dimensional parameter spaces with non-Euclidean structure. Information geometry frames this landscape using the Fisher information metric, enabling more prin…
Rethinking Disparity: A Depth Range Free Multi-View Stereo Based on Disparity
Existing learning-based multi-view stereo (MVS) methods rely on the depth range to build the 3D cost volume and may fail when the range is too large or unreliable. To address this problem, we propose a disparity-based MV…
GPUFrom Reward-Free Representations to Preferences: Rethinking Offline Preference-Based Reinforcement Learning
Preference-based reinforcement learning (PbRL) avoids explicit reward engineering by learning from pairwise human preference feedback. Existing offline PbRL methods typically follow a two-stage pipeline, first learning a…
Representation LearningReinforcement LearningOffline RLRethinking Direct Preference Optimization in Diffusion Models
Aligning text-to-image (T2I) diffusion models with human preferences has emerged as a critical research challenge. While recent advances in this area have extended preference optimization techniques from large language m…