德克薩斯大學提出：One-stage目標檢測最強算法 ExtremeNet

前戲

最近出了不少論文，各類SOTA。好比（點擊可訪問）：

商湯等提出：統一多目標跟蹤框架

亞馬遜提出：用於人羣計數的尺度感知注意力網絡

今天頭條推送的是目前人臉檢測方向的SOTA論文：改進SRN人臉檢測算法。本文要介紹的是目前(2019-01-26) one-stage目標檢測中最強算法：ExtremeNet。

正文

《Bottom-up Object Detection by Grouping Extreme and Center Points》

arXiv: https://arxiv.org/abs/1901.08043

github: https://github.com/xingyizhou/ExtremeNet

做者團隊：UT Austin

注：2019年01月23日剛出爐的paper

Abstract：With the advent of deep learning, object detection drifted from a bottom-up to a top-down recognition problem. State of the art algorithms enumerate a near-exhaustive list of object locations and classify each into: object or not. In this paper, we show that bottom-up approaches still perform competitively. We detect four extreme points (top-most, left-most, bottom-most, right-most) and one center point of objects using a standard keypoint estimation network. We group the five keypoints into a bounding box if they are geometrically aligned. Object detection is then a purely appearance-based keypoint estimation problem, without region classification or implicit feature learning. The proposed method performs on-par with the state-of-the-art region based detection methods, with a bounding box AP of 43.2% on COCO test-dev. In addition, our estimated extreme points directly span a coarse octagonal mask, with a COCO Mask AP of 18.9%, much better than the Mask AP of vanilla bounding boxes. Extreme point guided segmentation further improves this to 34.6% Mask AP.

Illustration of our object detection method

Illustration of our framework

Illustration of our object detection method

基礎工做

Extreme and center points

Keypoint detection

CornerNet

Deep Extreme Cut

創新點

Center Grouping

Ghost box suppression

Edge aggregation

Extreme Instance Segmentation

實驗結果

ExtremeNet有多強，看下面的圖示就知道了，在COCO test-dev數據集上，mAP爲43.2，在one-stage detector中，排名第一。惋惜的是沒有給出時間上的對比，論文中只介紹說測試一幅圖像，耗時322ms（3.1 FPS）。

State-of-the-art comparison on COCO test-dev

想要了解最新最快最好的論文速遞、開源項目和乾貨資料，歡迎加入CVer學術交流羣。涉及圖像分類、目標檢測、圖像分割、人臉檢測&識別、目標跟蹤、GANs、學術競賽交流、Re-ID、風格遷移、醫學影像分析、姿態估計、OCR、SLAM、場景文字檢測&識別和超分辨率等方向。