# Incorporating Depth into both CNN and CRF for Indoor Semantic Segmentation

arXiv: https://arxiv.org/abs/1705.07383 · HTML: https://mlsanity.com/paper/1705.07383/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 May 2017.

## Built on by (most cited first, 4 of 4)

- [Deep Learning Based 3D Segmentation: A Survey](https://mlsanity.com/paper/2103.05423/) (2021)
- [Striking the Right Balance: Recall Loss for Semantic Segmentation](https://mlsanity.com/paper/2106.14917/) (2021)
- [Deep Learning-Based 3D Instance and Semantic Segmentation: A Review](https://mlsanity.com/paper/2406.13308/) (2024)
- [Generating 3D Bio-Printable Patches Using Wound Segmentation and Reconstruction to Treat Diabetic Foot Ulcers](https://mlsanity.com/paper/2203.03814/) (2022)

## Abstract (arXiv)

To improve segmentation performance, a novel neural network architecture (termed DFCN-DCRF) is proposed, which combines an RGB-D fully convolutional neural network (DFCN) with a depth-sensitive fully-connected conditional random field (DCRF). First, a DFCN architecture which fuses depth information into the early layers and applies dilated convolution for later contextual reasoning is designed. Then, a depth-sensitive fully-connected conditional random field (DCRF) is proposed and combined with the previous DFCN to refine the preliminary result. Comparative experiments show that the proposed DFCN-DCRF has the best performance compared with most state-of-the-art methods.