# On Pre-Trained Image Features and Synthetic Images for Deep Learning

arXiv: https://arxiv.org/abs/1710.10710 · HTML: https://mlsanity.com/paper/1710.10710/

## Key facts

- Cited by 51 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 29 Oct 2017.

## Built on by (most cited first, 30 of 51)

- [Segment Anything in High Quality](https://mlsanity.com/paper/2306.01567/) (2023)
- [ZegCLIP: Towards Adapting CLIP for Zero-shot Semantic Segmentation](https://mlsanity.com/paper/2212.03588/) (2022)
- [BOP Challenge 2022 on Detection, Segmentation and Pose Estimation of Specific Rigid Objects](https://mlsanity.com/paper/2302.13075/) (2023)
- [Deep Regression on Manifolds: A 3D Rotation Case Study](https://mlsanity.com/paper/2103.16317/) (2021)
- [BOP Challenge 2023 on Detection, Segmentation and Pose Estimation of Seen and Unseen Rigid Objects](https://mlsanity.com/paper/2403.09799/) (2024)
- [X-Paste: Revisiting Scalable Copy-Paste for Instance Segmentation using CLIP and StableDiffusion](https://mlsanity.com/paper/2212.03863/) (2022)
- [MosaicFusion: Diffusion Models as Data Augmenters for Large Vocabulary Instance Segmentation](https://mlsanity.com/paper/2309.13042/) (2023)
- [DPODv2: Dense Correspondence-Based 6 DoF Pose Estimation](https://mlsanity.com/paper/2207.02805/) (2022)
- [Object Detection Using Sim2Real Domain Randomization for Robotic Applications](https://mlsanity.com/paper/2208.04171/) (2022)
- [Markerless Camera-to-Robot Pose Estimation via Self-supervised Sim-to-Real Transfer](https://mlsanity.com/paper/2302.14332/) (2023)
- [EgoEnv: Human-centric environment representations from egocentric video](https://mlsanity.com/paper/2207.11365/) (2022)
- [DALL-E for Detection: Language-driven Compositional Image Synthesis for Object Detection](https://mlsanity.com/paper/2206.09592/) (2022)
- [Tree Detection and Diameter Estimation Based on Deep Learning](https://mlsanity.com/paper/2210.17424/) (2022)
- [Enhancing Vision-Language Compositional Understanding with Multimodal Synthetic Data](https://mlsanity.com/paper/2503.01167/) (2025)
- [Generative Adversarial Networks for Image Super-Resolution: A Survey](https://mlsanity.com/paper/2204.13620/) (2022)
- [COPE: End-to-end trainable Constant Runtime Object Pose Estimation](https://mlsanity.com/paper/2208.08807/) (2022)
- [Beyond Generation: Harnessing Text to Image Models for Object Detection and Segmentation](https://mlsanity.com/paper/2309.05956/) (2023)
- [Evaluating a Synthetic Image Dataset Generated with Stable Diffusion](https://mlsanity.com/paper/2211.01777/) (2022)
- [Analysis of Training Object Detection Models with Synthetic Data](https://mlsanity.com/paper/2211.16066/) (2022)
- [Knock, knock. Who's there? -- Identifying football player jersey numbers with synthetic data](https://mlsanity.com/paper/2203.00734/) (2022)
- [A SAM-guided Two-stream Lightweight Model for Anomaly Detection](https://mlsanity.com/paper/2402.19145/) (2024)
- [Photo-realistic Neural Domain Randomization](https://mlsanity.com/paper/2210.12682/) (2022)
- [Object Detector Differences when using Synthetic and Real Training Data](https://mlsanity.com/paper/2312.00694/) (2023)
- [Federated Learning Over Images: Vertical Decompositions and Pre-Trained Backbones Are Difficult to Beat](https://mlsanity.com/paper/2309.03237/) (2023)
- [Label-Free Synthetic Pretraining of Object Detectors](https://mlsanity.com/paper/2208.04268/) (2022)
- [HaDR: Applying Domain Randomization for Generating Synthetic Multimodal Dataset for Hand Instance Segmentation in Cluttered Industrial Environments](https://mlsanity.com/paper/2304.05826/) (2023)
- [Harnessing Machine Learning for Discerning AI-Generated Synthetic Images](https://mlsanity.com/paper/2401.07358/) (2024)
- [Transfer learning with generative models for object detection on limited datasets](https://mlsanity.com/paper/2402.06784/) (2024)
- [How Important are Data Augmentations to Close the Domain Gap for Object Detection in Orbit?](https://mlsanity.com/paper/2410.15766/) (2024)
- [A Pipeline for Vision-Based On-Orbit Proximity Operations Using Deep Learning and Synthetic Imagery](https://mlsanity.com/paper/2101.05661/) (2021)

## Abstract (arXiv)

Deep Learning methods usually require huge amounts of training data to perform at their full potential, and often require expensive manual labeling. Using synthetic images is therefore very attractive to train object detectors, as the labeling comes for free, and several approaches have been proposed to combine synthetic and real images for training. In this paper, we show that a simple trick is sufficient to train very effectively modern object detectors with synthetic images only: We freeze the layers responsible for feature extraction to generic layers pre-trained on real images, and train only the remaining layers with plain OpenGL rendering. Our experiments with very recent deep architectures for object recognition (Faster-RCNN, R-FCN, Mask-RCNN) and image feature extractors (InceptionResnet and Resnet) show this simple approach performs surprisingly well.