Focal Loss for Dense Object Detection (Paper Summary)

Free AI web copilot to create summaries, insights and extended knowledge, download it at here

1299

Abstract

with the focal loss, RetinaNet is able to match the speed of previous one-stage detectors while surpassing the accuracy of all existing state-of-the-art two-stage detectors. Code is at: <a href="https://github.com/facebookresearch/Detectron.">https://github.com/facebookresearch/Detectron.</a></blockquote><blockquote id="aa0d">we present a one stage object detector that, for the first time, matches the state-of-the-art COCO AP of more complex two-stage detectors, such as the Feature Pyramid Network (FPN) [20] or Mask R-CNN [14] variants of Faster R-CNN [28]. To achieve this result, we identify class imbalance during training as the main obstacle impeding one-stage detector from achieving state-of-the-art accuracy and propose a new loss function that eliminates this barrier</blockquote><blockquote id="a999">A one-stage detector must process a much larger set of candidate object locations regularly sampled across an image. In practice this often amounts to enumerating ∼100k locations that densely cover spatial positions, scales, and aspect ratios. While similar sampling heuristics may also be applied, they are inefficient as the training procedure is still dominated by easily classified background examples. This inefficiency is a classic problem in object detection th

Options

at is typically addressed via techniques such as bootstrapping [33, 29] or hard example mining [37, 8, 31].</blockquote><blockquote id="b470">In this paper, we propose a new loss function that acts as a more effective alternative to previous approaches for dealing with class imbalance. The loss function is a dynamically scaled cross entropy loss, where the scaling factor decays to zero as confidence in the correct class increases, see Figure 1. Intuitively, this scaling factor can automatically down-weight the contribution of easy examples during training and rapidly focus the model on hard examples. Experiments show that our proposed Focal Loss enables us to train a high-accuracy, one-stage detector that significantly outperforms the alternatives of training with the sampling heuristics or hard example mining</blockquote><figure id="cea3"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*VPwNdzLC4gzQtI8De6P5HA.png"><figcaption></figcaption></figure><figure id="3d2b"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*WQnrRx2y-ybLmYNxCrcCCg.png"><figcaption></figcaption></figure><figure id="9863"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*AMqPCDdJT7DmMkVBsDXGEw.png"><figcaption></figcaption></figure></article></body>

Focal Loss for Dense Object Detection (Paper Summary)

Key insights:

The highest accuracy object detectors to date are based on a two-stage approach popularized by R-CNN, where a classifier is applied to a sparse set of candidate object locations. In contrast, one-stage detectors that are applied over a regular, dense sampling of possible object locations have the potential to be faster and simpler, but have trailed the accuracy of two-stage detectors thus far. In this paper, we investigate why this is the case. We discover that the extreme foreground-background class imbalance encountered during training of dense detectors is the central cause. We propose to address this class imbalance by reshaping the standard cross entropy loss such that it down-weights the loss assigned to well-classified examples. Our novel Focal Loss focuses training on a sparse set of hard examples and prevents the vast number of easy negatives from overwhelming the detector during training. To evaluate the effectiveness of our loss, we design and train a simple dense detector we call RetinaNet. Our results show that when trained with the focal loss, RetinaNet is able to match the speed of previous one-stage detectors while surpassing the accuracy of all existing state-of-the-art two-stage detectors. Code is at: https://github.com/facebookresearch/Detectron.

we present a one stage object detector that, for the first time, matches the state-of-the-art COCO AP of more complex two-stage detectors, such as the Feature Pyramid Network (FPN) [20] or Mask R-CNN [14] variants of Faster R-CNN [28]. To achieve this result, we identify class imbalance during training as the main obstacle impeding one-stage detector from achieving state-of-the-art accuracy and propose a new loss function that eliminates this barrier

A one-stage detector must process a much larger set of candidate object locations regularly sampled across an image. In practice this often amounts to enumerating ∼100k locations that densely cover spatial positions, scales, and aspect ratios. While similar sampling heuristics may also be applied, they are inefficient as the training procedure is still dominated by easily classified background examples. This inefficiency is a classic problem in object detection that is typically addressed via techniques such as bootstrapping [33, 29] or hard example mining [37, 8, 31].

In this paper, we propose a new loss function that acts as a more effective alternative to previous approaches for dealing with class imbalance. The loss function is a dynamically scaled cross entropy loss, where the scaling factor decays to zero as confidence in the correct class increases, see Figure 1. Intuitively, this scaling factor can automatically down-weight the contribution of easy examples during training and rapidly focus the model on hard examples. Experiments show that our proposed Focal Loss enables us to train a high-accuracy, one-stage detector that significantly outperforms the alternatives of training with the sampling heuristics or hard example mining