A Local Top-Down Module for Object Detection with Multi-scale Features

IvaNet: Learning to jointly detect and segment objets with the help of Local Top-Down Modules

“…by larger objects (persons) gradually as the receptive field becomes larger, which means the information backward from some higher layers is meaningless for some lower layers that are used to detect and segment small objects. To this end, a local top-down module (LTD) [4] has been proposed and demonstrated its functionality for single shot object detector [5] recently.…”

“…Object detection aims at locating objects with bounding boxes and classifying them into corresponding class. The exist-ing object detectors can be categorized into two branches, R-CNN based [8,9,10] and SSD based [5,1,4]. R-CNN based methods use a third part method to pre-select anchors, for instance, R-CNN [8] use the Selective Search [11].…”

“…There are some variants of SSD for improving the robustness by introducing the abstract semantic information from higher layers into lower ones via a top-down network. Instead using a full topdown module that will introduce much useless information as DSSD [1], LTD-SSD [4] proposes a local top-down module and achieves better results. Specifically, each prediction layer is integrated only with the upsampled features from its two succeeding layers.…”

“…Specifically, the parameters after block4 all are dropped. Furthermore, the same Local top-down module (LTD) presented in LTD-SSD [4] is adopted to construct a top-down feature pyramid. We note that the output channel of eah LTD is restricted to 384 by 1 × 1 convolutional layer, which is different from LTD-SSD.…”