Bag of Freebies for Training Object Detection Neural Networks
arXiv:1902.04103
Abstract
Training heuristics greatly improve various image classification model accuracies~\cite{he2018bag}. Object detection models, however, have more complex neural network structures and optimization targets. The training strategies and pipelines dramatically vary among different models. In this works, we explore training tweaks that apply to various models including Faster R-CNN and YOLOv3. These tweaks do not change the model architectures, therefore, the inference costs remain the same. Our empirical results demonstrate that, however, these freebies can improve up to 5% absolute precision compared to state-of-the-art baselines.
References in corpus (1)
Cited by in corpus (17)
- YOLOX: Exceeding YOLO Series in 2021
- Learning Spatial Fusion for Single-Shot Object Detection
- Gaussian Differential Privacy
- Learn an Effective Lip Reading Model without Pains
- Artificial and beneficial -- Exploiting artificial images for aerial vehicle detection
- Dynamic Mini-batch SGD for Elastic Distributed Training: Learning in the Limbo of Resources
- Data Augmentation for Object Detection via Differentiable Neural Rendering
- Scale-aware Automatic Augmentation for Object Detection
- IDM: An Intermediate Domain Module for Domain Adaptive Person Re-ID
- A unified neural network for object detection, multiple object tracking and vehicle re-identification
- DuBox: No-Prior Box Objection Detection via Residual Dual Scale Detectors
- Robust and Accurate Object Detection via Adversarial Learning
- GuidedMix-Net: Learning to Improve Pseudo Masks Using Labeled Images as Reference
- Mixup Regularization for Region Proposal based Object Detectors
- Towards Better Object Detection in Scale Variation with Adaptive Feature Selection
- DHARI Report to EPIC-Kitchens 2020 Object Detection Challenge
- Colonoscopy Polyp Detection: Domain Adaptation From Medical Report Images to Real-time Videos