# Training-Time-Friendly Network for Real-Time Object Detection

arXiv: https://arxiv.org/abs/1909.00700 · HTML: https://mlsanity.com/paper/1909.00700/

## Key facts

- Cited by 17 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 2 Sep 2019.

## Built on by (most cited first, 17 of 17)

- [MSG-Transformer: Exchanging Local Spatial Information by Manipulating Messenger Tokens](https://mlsanity.com/paper/2105.15168/) (2021)
- [DynamicISP: Dynamically Controlled Image Signal Processor for Image Recognition](https://mlsanity.com/paper/2211.01146/) (2022)
- [Rawgment: Noise-Accounted RAW Augmentation Enables Recognition in a Wide Variety of Environments](https://mlsanity.com/paper/2210.16046/) (2022)
- [Beyond Self-Supervision: A Simple Yet Effective Network Distillation Alternative to Improve Backbones](https://mlsanity.com/paper/2103.05959/) (2021)
- [Rigidity-Aware Detection for 6D Object Pose Estimation](https://mlsanity.com/paper/2303.12396/) (2023)
- [A Comprehensive Study of Real-Time Object Detection Networks Across Multiple Domains: A Survey](https://mlsanity.com/paper/2208.10895/) (2022)
- [ImplantFormer: Vision Transformer based Implant Position Regression Using Dental CBCT Data](https://mlsanity.com/paper/2210.16467/) (2022)
- [Stereo CenterNet based 3D Object Detection for Autonomous Driving](https://mlsanity.com/paper/2103.11071/) (2021)
- [Voxelized 3D Feature Aggregation for Multiview Detection](https://mlsanity.com/paper/2112.03471/) (2021)
- [PAFNet: An Efficient Anchor-Free Object Detector Guidance](https://mlsanity.com/paper/2104.13534/) (2021)
- [Avoiding Post-Processing with Event-Based Detection in Biomedical Signals](https://mlsanity.com/paper/2209.11007/) (2022)
- [Suppress-and-Refine Framework for End-to-End 3D Object Detection](https://mlsanity.com/paper/2103.10042/) (2021)
- [SAR-ShipNet: SAR-Ship Detection Neural Network via Bidirectional Coordinate Attention and Multi-resolution Feature Fusion](https://mlsanity.com/paper/2203.15480/) (2022)
- [A Coarse to Fine Framework for Object Detection in High Resolution Image](https://mlsanity.com/paper/2303.01219/) (2023)
- [Streaming Object Detection on Fisheye Cameras for Automatic Parking](https://mlsanity.com/paper/2305.14713/) (2023)
- [Rethinking the Detection Head Configuration for Traffic Object Detection](https://mlsanity.com/paper/2210.03883/) (2022)
- [On the Importance of Large Objects in CNN Based Object Detection Algorithms](https://mlsanity.com/paper/2311.11714/) (2023)

## Abstract (arXiv)

Modern object detectors can rarely achieve short training time, fast inference speed, and high accuracy at the same time. To strike a balance among them, we propose the Training-Time-Friendly Network (TTFNet). In this work, we start with light-head, single-stage, and anchor-free designs, which enable fast inference speed. Then, we focus on shortening training time. We notice that encoding more training samples from annotated boxes plays a similar role as increasing batch size, which helps enlarge the learning rate and accelerate the training process. To this end, we introduce a novel approach using Gaussian kernels to encode training samples. Besides, we design the initiative sample weights for better information utilization. Experiments on MS COCO show that our TTFNet has great advantages in balancing training time, inference speed, and accuracy. It has reduced training time by more than seven times compared to previous real-time detectors while maintaining state-of-the-art performances. In addition, our super-fast version of TTFNet-18 and TTFNet-53 can outperform SSD300 and YOLOv3 by less than one-tenth of their training time, respectively. The code has been made available at urlhttps://github.com/ZJULearning/ttfnet.