Fast Vehicle and Pedestrian Detection Using Improved Mask R-CNN

Figure 2

Six network structures for feature extraction. (a) Featurized image pyramid. (b) Single feature map. (c) Pyramidal feature hierarchy. (d) Top-down pyramid. (e) Finest level only. (f) Feature pyramid network.

(a)

(b)

(c)

(d)

(e)

(f)