A Meta-Theory of Boundary Detection Benchmarks
This work addresses the issue of benchmark reliability for researchers in computer vision and boundary detection, but it is incremental as it focuses on refining existing evaluation methods rather than introducing a new paradigm.
The paper tackles the problem of unreliable human-labeled benchmarks in boundary detection by proposing a computational framework to remove inappropriate labels and estimate intrinsic boundary properties, resulting in a method to improve evaluation accuracy.
Human labeled datasets, along with their corresponding evaluation algorithms, play an important role in boundary detection. We here present a psychophysical experiment that addresses the reliability of such benchmarks. To find better remedies to evaluate the performance of any boundary detection algorithm, we propose a computational framework to remove inappropriate human labels and estimate the intrinsic properties of boundaries.