# Learning from THEODORE: A Synthetic Omnidirectional Top-View Indoor Dataset for Deep Transfer Learning

arXiv: https://arxiv.org/abs/2011.05719 · HTML: https://mlsanity.com/paper/2011.05719/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 11 Nov 2020.

## Built on by (most cited first, 9 of 9)

- [Surround-view Fisheye Camera Perception for Automated Driving: Overview, Survey and Challenges](https://mlsanity.com/paper/2205.13281/) (2022)
- [OmniFlow: Human Omnidirectional Optical Flow](https://mlsanity.com/paper/2104.07960/) (2021)
- [ADD: An Automatic Desensitization Fisheye Dataset for Autonomous Driving](https://mlsanity.com/paper/2308.07590/) (2023)
- [Human Pose Estimation in Monocular Omnidirectional Top-View Images](https://mlsanity.com/paper/2304.08186/) (2023)
- [Applications of Deep Learning for Top-View Omnidirectional Imaging: A Survey](https://mlsanity.com/paper/2304.08193/) (2023)
- [Rethinking Camera Choice: An Empirical Study on Fisheye Camera Properties in Robotic Manipulation](https://mlsanity.com/paper/2603.02139/) (2026)
- [Sector Patch Embedding: An Embedding Module Conforming to The Distortion Pattern of Fisheye Image](https://mlsanity.com/paper/2303.14645/) (2023)
- [Towards Adaptive Fusion of Multimodal Deep Networks for Human Action Recognition](https://mlsanity.com/paper/2512.04943/) (2025)
- [Geometric Domain Adaptation via Optimal Transport for Linear Regression in R^2](https://mlsanity.com/paper/2606.14023/) (2026)

## Abstract (arXiv)

Recent work about synthetic indoor datasets from perspective views has shown significant improvements of object detection results with Convolutional Neural Networks(CNNs). In this paper, we introduce THEODORE: a novel, large-scale indoor dataset containing 100,000 high-resolution diversified fisheye images with 14 classes. To this end, we create 3D virtual environments of living rooms, different human characters and interior textures. Beside capturing fisheye images from virtual environments we create annotations for semantic segmentation, instance masks and bounding boxes for object detection tasks. We compare our synthetic dataset to state of the art real-world datasets for omnidirectional images. Based on MS COCO weights, we show that our dataset is well suited for fine-tuning CNNs for object detection. Through a high generalization of our models by means of image synthesis and domain randomization, we reach an AP up to 0.84 for class person on High-Definition Analytics dataset.