Human-In-The-Loop Document Layout Analysis
Not on the map. Why?
Abstract (arXiv)
Document layout analysis (DLA) aims to divide a document image into different types of regions. DLA plays an important role in the document content understanding and information extraction systems. Exploring a method that can use less data for effective training contributes to the development of DLA. We consider a Human-in-the-loop (HITL) collaborative intelligence in the DLA. Our approach was inspired by the fact that the HITL push the model to learn from the unknown problems by adding a small amount of data based on knowledge. The HITL select key samples by using confidence. However, using confidence to find key samples is not suitable for DLA tasks. We propose the Key Samples Selection (KSS) method to find key samples in high-level tasks (semantic segmentation) more accurately through agent collaboration, effectively reducing costs. Once selected, these key samples are passed to human beings for active labeling, then the model will be updated with the labeled samples. Hence, we revisited the learning system from reinforcement learning and designed a sample-based agent update strategy, which effectively improves the agent's ability to accept new samples. It achieves significant improvement results in two benchmarks (DSSE-200 (from 77.1% to 86.3%) and CS-150 (from 88.0% to 95.6%)) by using 10% of labeled data.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 27 of its 74 references that are in the map. First posted on arXiv 4 Aug 2021.
Builds on
27 of 74 references.
References · 27
- 2015Deep Residual Learning for Image Recognitioncited by 52,096
- 2014Microsoft COCO: Common Objects in Contextcited by 17,608
- 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
- 2014Fully Convolutional Networks for Semantic Segmentationcited by 4,859
- 2016DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFscited by 3,241
- 2017Prototypical Networks for Few-shot Learningcited by 3,015
- 2016Pyramid Scene Parsing Networkcited by 2,479
- 2016Xception: Deep Learning with Depthwise Separable Convolutionscited by 2,332
- 2016Matching Networks for One Shot Learningcited by 2,255
- 2015SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentationcited by 1,487
- 2015LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loopcited by 1,355
- 2016Learning to learn by gradient descent by gradient descentcited by 620
15 more
- 2019LayoutLM: Pre-training of Text and Layout for Document Image Understandingcited by 553
- 2018Meta-Learning for Semi-Supervised Few-Shot Classificationcited by 456
- 2018Dynamic Few-Shot Visual Learning without Forgettingcited by 385
- 2020Differentiable Augmentation for Data-Efficient GAN Trainingcited by 363
- 2018BabyAI: A Platform to Study the Sample Efficiency of Grounded Language Learningcited by 244
- 2020DocBank: A Benchmark Dataset for Document Layout Analysiscited by 153
- 2018Pyramid Attention Network for Semantic Segmentationcited by 149
- 2020Interventional Few-Shot Learningcited by 112
- 2019Learning to Predict Layout-to-image Conditional Convolutions for Semantic Image Synthesiscited by 108
- 2017Learning to Extract Semantic Structure from Documents Using Multimodal Fully Convolutional Neural Networkcited by 92
- 2019Graph Few-shot Learning via Knowledge Transfercited by 63
- 2014Interactive Policy Learning through Confidence-Based Autonomycited by 29
- 2020Cross-Domain Document Object Detection: Benchmark Suite and Methodcited by 24
- 2017Fully Convolutional Neural Networks for Page Segmentation of Historical Document Imagescited by 9
- 2021Document Layout Analysis via Dynamic Residual Feature Fusioncited by 4
Built on by
Citing papers per year: 2022: 1; 2023: 1.
2 papers.