ICCV-2017
1 Background and Motivation
NMS 是许多目标检测算法重要的步骤之一
NMS 的缺点,if an object lies within the predefined overlap threshold, it
leads to a miss.(sets the score for neighboring detections to zero)
Intuitively, if a bounding box has a very high overlap with
M
M
M(maximum score), it should be assigned a very low score, while if it has a low overlap, it can maintain its original detection score.
作者对 NMS 进行了改进
2 Related Work
NMS first employed in edge detection techniques
3 Advantages / Contributions
- NMS 基础上提出 Soft-NMS,not require any extra training and is simple to implement
- obtain consistent improvements for the coco-style mAP metric on
standard datasets like PASCALVOC2007 (1.7% for both R-FCN and Faster-RCNN) and MS-COCO
4 Method
NMS 与 maximum score box 的 IoU 大于阈值的都抑制为了 0
全抑制为0,容易漏检,NMS should take the following conditions into account,
NMS 的阈值越高,越容易误检,越低约容易漏检
高 NMS 阈值情况下,mAP 反而会下降,因为 the increase in false positives would be much higher than the increase in true positives for this case because the number of objects is typically much smaller than the number of RoIs generated by a detector.
soft-NMS 大于阈值的分数降低
a list of detection boxes
B
B
B with scores
S
S
S.
maximum score
M
M
M
the set of final detections
D
D
D
suppressing all nearby detection boxes with a low N t N_t Nt would increase the miss-rate.
其中 Rescoring Functions f ( i o u ( M , b i ) ) f(iou(M, b_i)) f(iou(M,bi)) 的形式有如下两种
(1)非连续的(not continuous)
与最高分box 的 IoU 越大,分数降得更多(误检的概率比较大),IoU 越小,分数降的越少
IoU 等于 N t N_t Nt 的时候,值不一样
(2)连续的
计算复杂度同 traditional NMS
also a greedy algorithm and does not find the globally optimal re-scoring of detection boxes.
不一定局限于指数形式,也可以是 Gompertz function
5 Experiments
5.1 Results
N
t
N_t
Nt to 0.3,
σ
\sigma
σ to 0.5
Soft-NMS has more potential to improve recall at higher O t O_t Ot(detection evaluation threshold).
5.2 Sensitivity Analysis
试试 soft-NMS 超参数 σ \sigma σ 的敏感性
5.3 When does Soft-NMS work better?
(1)Localization Performance
千言万语,不如下面一个表来的直接
even when we obtain better performance at higher
O
t
O_t
Ot, performance at lower
O
t
O_t
Ot does not drop.
low values of σ \sigma σ perform better at lower O t O_t Ot and higher values of sigma perform better at higher O t O_t Ot.
(2)Precision vs Recall
5.4 Qualitative Results
a large wide bounding box spanning multiple people is suppressed because it had a small overlap with multiple detection boxes with a higher score than it.
文章来源:https://www.toymoban.com/news/detail-600326.html
6 Conclusion(own)
soft-NMS也是一种贪心算法,并不能保证找到全局最优的检测框分数重置。文章来源地址https://www.toymoban.com/news/detail-600326.html
到了这里,关于【Soft NMS】《Soft-NMS – Improving Object Detection With One Line of Code》的文章就介绍完了。如果您还想了解更多内容,请在右上角搜索TOY模板网以前的文章或继续浏览下面的相关文章,希望大家以后多多支持TOY模板网!