[论文解读] On the Utility of Context (or the Lack Thereof) for Object Detection.

Ehud Barnea, Ohad Ben‐Shahar|arXiv (Cornell University)|Nov 15, 2017

Advanced Image and Video Retrieval Techniques被引用 4

一句话总结

本文通过将检测分数建模为检测器输出与上下文关系的函数，将上下文利用视为一个优化问题，研究了为何上下文通常仅带来微小的性能提升。研究发现，上下文主要有助于纠正分类错误，但无法缓解定位错误，这解释了其在实际检测性能中改进有限的原因。

ABSTRACT

The recurring context in which objects appear holds valuable information that can be employed to predict their existence. This intuitive observation indeed led many researchers to endow appearance-based detectors with explicit reasoning about context. The underlying thesis suggests that with stronger contextual relations, the better improvement in detection capacity one can expect from such a combined approach. In practice, however, the observed improvement in many case is modest at best, and often only marginal. In this work we seek to understand this phenomenon better, in part by pursuing an opposite approach. Instead of going from context to detection score, we formulate the score as a function of standard detector results and contextual relations, an approach that allows to treat the utility of context as an optimization problem in order to obtain the largest gain possible from considering context in the first place. Analyzing different contextual relations reveals the most helpful ones and shows that in many cases including context can help while in other cases a significant improvement is simply impossible or impractical. To better understand these results we then analyze the ability of context to handle different types of false detections, revealing that contextual information cannot ameliorate localization errors, which in turn also diminish the observed improvement obtained by correcting other types of errors. These insights provide further explanations and better understanding regarding the success or failure of utilizing context for object detection.

研究动机与目标

• 理解为何在目标检测中引入上下文在实践中仅带来适度的性能提升。
• 研究上下文推理在纠正不同类型检测错误方面的局限性。
• 将上下文的效用建模为一个优化问题，以最大化检测性能的提升。
• 识别出对提升检测性能最有效的上下文关系类型。
• 确定在何种情况下引入上下文在实践上不可行或无法使检测性能超过某一阈值。

提出的方法

• 将检测分数表述为标准检测器输出与上下文关系的可学习函数，从而实现对上下文贡献的优化。
• 分析各种类型的上下文关系（如空间关系、语义关系）以评估其对检测性能的相对影响。
• 将上下文集成视为一个优化问题，以识别上下文信息可能带来的最大性能增益。
• 评估上下文纠正不同类型错误（尤其是误报和定位错误）的能力。
• 通过实证分析比较在不同错误类别下，有无上下文时的检测性能。
• 应用该框架识别出上下文无效的情境，特别是当定位错误占主导地位时。

实验结果

研究问题

• RQ1 为何在目标检测中集成上下文通常仅带来微小的性能提升？
• RQ2 哪些类型的上下文关系能带来最显著的检测准确率增益？
• RQ3 上下文信息能否有效纠正目标检测中的定位错误？
• RQ4 在何种条件下，使用上下文无法或不切实际地提升检测性能？
• RQ5 上下文的效用在不同类型误检（如误报与定位错误）之间如何变化？

主要发现

• 上下文信息无法纠正定位错误，这严重限制了其在提升检测性能方面的整体有效性。
• 上下文的主要优势在于减少误报，尤其是通过语义和空间推理实现。
• 即使实现最优的上下文集成，性能增益仍受定位错误存在的制约，因为上下文无法解决此类错误。
• 某些错误类型（尤其是涉及不准确边界框预测的错误）会使基于上下文的纠正方法失效。
• 由于定位错误具有不可逆性，上下文的理论潜力在实践中往往无法实现。
• 本研究发现，上下文的效用从根本上受限于基础检测器输出中存在的错误类型。