An Implementation of Faster RCNN with Study for Region Sampling

02/07/2017
by   Xinlei Chen, et al.
0

We adapted the join-training scheme of Faster RCNN framework from Caffe to TensorFlow as a baseline implementation for object detection. Our code is made publicly available. This report documents the simplifications made to the original pipeline, with justifications from ablation analysis on both PASCAL VOC 2007 and COCO 2014. We further investigated the role of non-maximal suppression (NMS) in selecting regions-of-interest (RoIs) for region classification, and found that a biased sampling toward small regions helps performance and can achieve on-par mAP to NMS-based sampling when converged sufficiently.

READ FULL TEXT

page 1

page 2

page 3

research
09/02/2020

Lifelong Object Detection

Recent advances in object detection have benefited significantly from ra...
research
11/24/2017

Feature Selective Networks for Object Detection

Objects for detection usually have distinct characteristics in different...
research
12/25/2016

YOLO9000: Better, Faster, Stronger

We introduce YOLO9000, a state-of-the-art, real-time object detection sy...
research
10/02/2020

DIRV: Dense Interaction Region Voting for End-to-End Human-Object Interaction Detection

Recent years, human-object interaction (HOI) detection has achieved impr...
research
02/09/2022

Point-Level Region Contrast for Object Detection Pre-Training

In this work we present point-level region contrast, a self-supervised p...
research
03/30/2017

DeNet: Scalable Real-time Object Detection with Directed Sparse Sampling

We define the object detection from imagery problem as estimating a very...
research
07/23/2022

SSBNet: Improving Visual Recognition Efficiency by Adaptive Sampling

Downsampling is widely adopted to achieve a good trade-off between accur...

Please sign up or login with your details

Forgot password? Click here to reset