# Learning Fully Dense Neural Networks for Image Semantic Segmentation

arXiv: https://arxiv.org/abs/1905.08929 · HTML: https://mlsanity.com/paper/1905.08929/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 22 May 2019.

## Built on by (most cited first, 9 of 9)

- [Transferable Semantic Augmentation for Domain Adaptation](https://mlsanity.com/paper/2103.12562/) (2021)
- [Automatic universal taxonomies for multi-domain semantic segmentation](https://mlsanity.com/paper/2207.08445/) (2022)
- [BEFD: Boundary Enhancement and Feature Denoising for Vessel Segmentation](https://mlsanity.com/paper/2104.03768/) (2021)
- [Weakly supervised training of universal visual concepts for multi-domain semantic segmentation](https://mlsanity.com/paper/2212.10340/) (2022)
- [Spatial-information Guided Adaptive Context-aware Network for Efficient RGB-D Semantic Segmentation](https://mlsanity.com/paper/2308.06024/) (2023)
- [A Simple and Generalist Approach for Panoptic Segmentation](https://mlsanity.com/paper/2408.16504/) (2024)
- [Use square root affinity to regress labels in semantic segmentation](https://mlsanity.com/paper/2103.04990/) (2021)
- [RL-CoSeg : A Novel Image Co-Segmentation Algorithm with Deep Reinforcement Learning](https://mlsanity.com/paper/2204.05951/) (2022)
- [Efficient RGB-D Scene Understanding via Multi-task Adaptive Learning and Cross-dimensional Feature Guidance](https://mlsanity.com/paper/2603.07570/) (2026)

## Abstract (arXiv)

Semantic segmentation is pixel-wise classification which retains critical spatial information. The "feature map reuse" has been commonly adopted in CNN based approaches to take advantage of feature maps in the early layers for the later spatial reconstruction. Along this direction, we go a step further by proposing a fully dense neural network with an encoder-decoder structure that we abbreviate as FDNet. For each stage in the decoder module, feature maps of all the previous blocks are adaptively aggregated to feed-forward as input. On the one hand, it reconstructs the spatial boundaries accurately. On the other hand, it learns more efficiently with the more efficient gradient backpropagation. In addition, we propose the boundary-aware loss function to focus more attention on the pixels near the boundary, which boosts the "hard examples" labeling. We have demonstrated the best performance of the FDNet on the two benchmark datasets: PASCAL VOC 2012, NYUDv2 over previous works when not considering training on other datasets.