# CRF Learning with CNN Features for Image Segmentation

arXiv: https://arxiv.org/abs/1503.08263 · HTML: https://mlsanity.com/paper/1503.08263/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 28 Mar 2015.

## Built on by (most cited first, 6 of 6)

- [Semantic Probabilistic Layers for Neuro-Symbolic Learning](https://mlsanity.com/paper/2206.00426/) (2022)
- [Introspective Robot Perception using Smoothed Predictions from Bayesian Neural Networks](https://mlsanity.com/paper/2109.12869/) (2021)
- [Joint Learning of Frequency and Spatial Domains for Dense Predictions](https://mlsanity.com/paper/2202.08991/) (2022)
- [Model-based inexact graph matching on top of CNNs for semantic scene understanding](https://mlsanity.com/paper/2301.07468/) (2023)
- [Enhancing Lossy Compression Through Cross-Field Information for Scientific Applications](https://mlsanity.com/paper/2409.18295/) (2024)
- [Predicting Road Crossing Behaviour using Pose Detection and Sequence Modelling](https://mlsanity.com/paper/2508.15336/) (2025)

## Abstract (arXiv)

Conditional Random Rields (CRF) have been widely applied in image segmentations. While most studies rely on hand-crafted features, we here propose to exploit a pre-trained large convolutional neural network (CNN) to generate deep features for CRF learning. The deep CNN is trained on the ImageNet dataset and transferred to image segmentations here for constructing potentials of superpixels. Then the CRF parameters are learnt using a structured support vector machine (SSVM). To fully exploit context information in inference, we construct spatially related co-occurrence pairwise potentials and incorporate them into the energy function. This prefers labelling of object pairs that frequently co-occur in a certain spatial layout and at the same time avoids implausible labellings during the inference. Extensive experiments on binary and multi-class segmentation benchmarks demonstrate the promise of the proposed method. We thus provide new baselines for the segmentation performance on the Weizmann horse, Graz-02, MSRC-21, Stanford Background and PASCAL VOC 2011 datasets.