---
title: "FDSML: Formula-Driven Supervised Metric Learning with Parameter-Aware Triplet Loss"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2610.39423v1"
permalink: "https://feed.astralor.com/items/itm_3dfdc93d34eb1ff0a01bd2e8"
published: "2026-10-29T06:02:45.000Z"
score: 54
channel: "ai"
tags: ["RAG","多模态","FDSML","Formula-Driven Supervised Metric","Learning","Parameter-Aware Triplet Loss","Formula-Driven Supervised Learning","FDSL"]
---

# FDSML: Formula-Driven Supervised Metric Learning with Parameter-Aware Triplet Loss

## 速览

Formula-Driven Supervised Learning (FDSL) enables vision pre-training without real images, but existing methods cast pre-training as classification on discretized labels, discarding continuity in the generative process.…

## 关键点

- Formula-Driven Supervised Learning (FDSL) enables vision pre-training without real images, but existing methods cast pre-training as classification on discretized labels, discarding continuity in the generative process.
- We propose Formula-Driven Supervised Metric Learning (FDSML), which derives relational supervision from image-generation parameters beyond discrete class labels.
- FDSML combines FineFractalDB, our refined fractal dataset with higher intra-class variation and more stable structures, with a Parameter-Aware Triplet (PA-Triplet) Loss that assigns positives and negatives according to …

## 全文

Formula-Driven Supervised Learning (FDSL) enables vision pre-training without real images, but existing methods cast pre-training as classification on discretized labels, discarding continuity in the generative process. We propose Formula-Driven Supervised Metric Learning (FDSML), which derives relational supervision from image-generation parameters beyond discrete class labels. FDSML combines FineFractalDB, our refined fractal dataset with higher intra-class variation and more stable structures, with a Parameter-Aware Triplet (PA-Triplet) Loss that assigns positives and negatives according to distances in the underlying Iterated Function System parameter space. This objective trains embeddings to respect parameter-derived relative similarity orderings. Experiments show that FDSML improves transferability to image retrieval while remaining competitive on downstream classification and improving robustness under several corruption settings. On ResNet-50, Recall@1 improves from 65.6% to 74.0% on Stanford Online Products and from 56.2% to 78.4% on Stanford Cars over FDSL pre-training. These results suggest that parameter-derived relational supervision provides a stronger synthetic pre-training signal than label-based classification for transfer beyond recognition.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_3dfdc93d34eb1ff0a01bd2e8