# Deep Value Networks Learn to Evaluate and Iteratively Refine Structured Outputs

arXiv: https://arxiv.org/abs/1703.04363 · HTML: https://mlsanity.com/paper/1703.04363/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 13 Mar 2017.

## Built on by (most cited first, 12 of 12)

- [Neuro-Symbolic Constraint Programming for Structured Prediction](https://mlsanity.com/paper/2103.17232/) (2021)
- [Vector-Valued Least-Squares Regression under Output Regularity Assumptions](https://mlsanity.com/paper/2211.08958/) (2022)
- [Sketch In, Sketch Out: Accelerating both Learning and Inference for Structured Prediction with Kernels](https://mlsanity.com/paper/2302.10128/) (2023)
- [Flow-based Spatio-Temporal Structured Prediction of Motion Dynamics](https://mlsanity.com/paper/2104.04391/) (2021)
- [Energy-Based Models for Cross-Modal Localization using Convolutional Transformers](https://mlsanity.com/paper/2306.04021/) (2023)
- [Structured Convolutional Kernel Networks for Airline Crew Scheduling](https://mlsanity.com/paper/2105.11646/) (2021)
- [Deep Sketched Output Kernel Regression for Structured Prediction](https://mlsanity.com/paper/2406.09253/) (2024)
- [Directly Training Joint Energy-Based Models for Conditional Synthesis and Calibrated Prediction of Multi-Attribute Data](https://mlsanity.com/paper/2108.04227/) (2021)
- [SOInter: A Novel Deep Energy Based Interpretation Method for Explaining Structured Output Models](https://mlsanity.com/paper/2202.09914/) (2022)
- [Structured Prediction in NLP -- A survey](https://mlsanity.com/paper/2110.02057/) (2021)
- [Implicit Training of Energy Model for Structure Prediction](https://mlsanity.com/paper/2211.11649/) (2022)
- [Interval Type-2 Fuzzy Neural Networks for Multi-Label Classification](https://mlsanity.com/paper/2302.10430/) (2023)

## Abstract (arXiv)

We approach structured output prediction by optimizing a deep value network (DVN) to precisely estimate the task loss on different output configurations for a given input. Once the model is trained, we perform inference by gradient descent on the continuous relaxations of the output variables to find outputs with promising scores from the value network. When applied to image segmentation, the value network takes an image and a segmentation mask as inputs and predicts a scalar estimating the intersection over union between the input and ground truth masks. For multi-label classification, the DVN's objective is to correctly predict the F1 score for any potential label configuration. The DVN framework achieves the state-of-the-art results on multi-label prediction and image segmentation benchmarks.