RECORD
High-for-Low and Low-for-High: Efficient Boundary Detection from Deep Object Features and its Applications to High-Level Vision
Computer Vision and Pattern Recognition · Submitted 23 Apr 2015
abstract compiled · 2026-09-23 · no language-model review
The authors' main claim Thus, our contributions include a boundary detection system that is accurate, efficient, generalizes well to multiple datasets, and is also shown to improve existing state-of-the-art high-level vision methods on three distinct tasks.
The compiled record
The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
Open the compiled record
The abstract, sentence by sentence
Most of the current boundary detection systems rely exclusively on low-level features, such as color and texture. However, perception studies suggest that humans employ object-level reasoning when judging if a particular pixel is a boundary. Inspired by this observation, in this work we show how to predict boundaries by exploiting object-level features from a pretrained object-classification network. Our method can be viewed as a "High-for-Low" approach where high-level object features inform the low-level boundary detection process. Our model achieves state-of-the-art performance on an established boundary detection benchmark and it is efficient to run. Additionally, we show that due to the semantic nature of our boundaries we can use them to aid a number of high-level vision tasks. We demonstrate that using our boundaries we improve the performance of state-of-the-art methods on the problems of semantic boundary labeling, semantic segmentation and object proposal generation. We can view this process as a "Low-for-High" scheme, where low-level boundaries aid high-level vision tasks. Thus, our contributions include a boundary detection system that is accurate, efficient, generalizes well to multiple datasets, and is also shown to improve existing state-of-the-art high-level vision methods on three distinct tasks.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Most of the current boundary detection systems rely exclusively on low-level features, such as color and texture. | background 0.98 | 0.05 | 0.02 | asserted 1.00 |
| s1 | However, perception studies suggest that humans employ object-level reasoning when judging if a particular pixel is a boundary. | background 0.91 | 0.04 | 0.03 | hedged possibility 0.94 |
| s2 | Inspired by this observation, in this work we show how to predict boundaries by exploiting object-level features from a pretrained object-classification network. | contribution 0.91 | 0.87 | 0.13 | asserted 1.00 |
| s3 | Our method can be viewed as a "High-for-Low" approach where high-level object features inform the low-level boundary detection process. | method 0.54 | 0.87 | 0.12 | asserted 0.94 |
| s4 | Our model achieves state-of-the-art performance on an established boundary detection benchmark and it is efficient to run. | result 1.00 | 0.96 | 0.26 | asserted 1.00 |
| s5 | Additionally, we show that due to the semantic nature of our boundaries we can use them to aid a number of high-level vision tasks. | interpretation 0.45 | 0.86 | 0.13 | asserted 0.90 |
| s6 | We demonstrate that using our boundaries we improve the performance of state-of-the-art methods on the problems of semantic boundary labeling, semantic segmentation and object proposal generation. | result 0.96 | 0.95 | 0.17 | asserted 1.00 |
| s7 | We can view this process as a "Low-for-High" scheme, where low-level boundaries aid high-level vision tasks. | interpretation 0.94 | 0.80 | 0.10 | asserted 0.79 |
| s8 | Thus, our contributions include a boundary detection system that is accurate, efficient, generalizes well to multiple datasets, and is also shown to improve existing state-of-the-art high-level vision methods on three distinct tasks. | contribution 1.00 | 0.94 | 0.73 | asserted 1.00 |
What was examined, and what was not
- arXiv listing abstract v3
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
The paper
Original abstract
Most of the current boundary detection systems rely exclusively on low-level features, such as color and texture. However, perception studies suggest that humans employ object-level reasoning when judging if a particular pixel is a boundary. Inspired by this observation, in this work we show how to predict boundaries by exploiting object-level features from a pretrained object-classification network. Our method can be viewed as a "High-for-Low" approach where high-level object features inform the low-level boundary detection process. Our model achieves state-of-the-art performance on an established boundary detection benchmark and it is efficient to run. Additionally, we show that due to the semantic nature of our boundaries we can use them to aid a number of high-level vision tasks. We demonstrate that using our boundaries we improve the performance of state-of-the-art methods on the problems of semantic boundary labeling, semantic segmentation and object proposal generation. We can view this process as a "Low-for-High" scheme, where low-level boundaries aid high-level vision tasks. Thus, our contributions include a boundary detection system that is accurate, efficient, generalizes well to multiple datasets, and is also shown to improve existing state-of-the-art high-level vision methods on three distinct tasks.
Human review
How the literature uses it
What this paper does with the arXiv papers it cites
What it cites
Uses the method or tool of 1 paper
-
We use the Caffe library [ref] to implement our network architecture.
Caffe: Convolutional Architecture for Fast Feature Embedding · Implementation Details · 1.00
Discussion
Know something about this paper? Leave a comment, or highlight any passage above to comment on it.