# Do We Need More Training Data?

arXiv: https://arxiv.org/abs/1503.01508 · HTML: https://mlsanity.com/paper/1503.01508/

## Key facts

- Cited by 27 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Mar 2015.

## Built on by (most cited first, 27 of 27)

- [Transfer Learning with Synthetic Corpora for Spatial Role Labeling and Reasoning](https://mlsanity.com/paper/2210.16952/) (2022)
- [Open Long-Tailed Recognition in a Dynamic World](https://mlsanity.com/paper/2208.08349/) (2022)
- [General Intelligence Requires Rethinking Exploration](https://mlsanity.com/paper/2211.07819/) (2022)
- [AugLiChem: Data Augmentation Library of Chemical Structures for Machine Learning](https://mlsanity.com/paper/2111.15112/) (2021)
- [Scaling up instance annotation via label propagation](https://mlsanity.com/paper/2110.02277/) (2021)
- [Multi-Dataset Co-Training with Sharpness-Aware Optimization for Audio Anti-spoofing](https://mlsanity.com/paper/2305.19953/) (2023)
- [Oriented object detection in optical remote sensing images using deep learning: a survey](https://mlsanity.com/paper/2302.10473/) (2023)
- [Quantifying Privacy Risks of Prompts in Visual Prompt Learning](https://mlsanity.com/paper/2310.11970/) (2023)
- [Deceptive Level Generation for Angry Birds](https://mlsanity.com/paper/2106.01639/) (2021)
- [VisionKG: Unleashing the Power of Visual Datasets via Knowledge Graph](https://mlsanity.com/paper/2309.13610/) (2023)
- [ApacheJIT: A Large Dataset for Just-In-Time Defect Prediction](https://mlsanity.com/paper/2203.00101/) (2022)
- [Long-tailed Recognition by Learning from Latent Categories](https://mlsanity.com/paper/2206.01010/) (2022)
- [A Pilot Study For Fragment Identification Using 2D NMR and Deep Learning](https://mlsanity.com/paper/2103.12169/) (2021)
- [Data Deletion for Linear Regression with Noisy SGD](https://mlsanity.com/paper/2410.09311/) (2024)
- [UtilGen: Utility-Centric Generative Data Augmentation with Dual-Level Task Adaptation](https://mlsanity.com/paper/2510.24262/) (2025)
- [Identifying High Accuracy Regions in Traffic Camera Images to Enhance the Estimation of Road Traffic Metrics: A Quadtree-Based Method](https://mlsanity.com/paper/2106.14049/) (2021)
- [Deep Convolutional Neural Networks for Onychomycosis Detection](https://mlsanity.com/paper/2106.16139/) (2021)
- [Accelerating the creation of instance segmentation training sets through bounding box annotation](https://mlsanity.com/paper/2205.11563/) (2022)
- [SATDAUG -- A Balanced and Augmented Dataset for Detecting Self-Admitted Technical Debt](https://mlsanity.com/paper/2403.07690/) (2024)
- [Collate: Collaborative Neural Network Learning for Latency-Critical Edge Systems](https://mlsanity.com/paper/2607.08013/) (2026)
- [Demystifying Ten Big Ideas and Rules Every Fire Scientist & Engineer Should Know About Blackbox, Whitebox & Causal Artificial Intelligence](https://mlsanity.com/paper/2111.13756/) (2021)
- [Trustworthiness of Laser-Induced Breakdown Spectroscopy Predictions via Simulation-based Synthetic Data Augmentation and Multitask Learning](https://mlsanity.com/paper/2210.03762/) (2022)
- [Designing deep neural networks for driver intention recognition](https://mlsanity.com/paper/2402.05150/) (2024)
- [Generative Active Learning with Variational Autoencoder for Radiology Data Generation in Veterinary Medicine](https://mlsanity.com/paper/2403.03642/) (2024)
- [Toward Robust Canine Cardiac Diagnosis: Deep Prototype Alignment Network-Based Few-Shot Segmentation in Veterinary Medicine](https://mlsanity.com/paper/2403.06471/) (2024)
- [CBNN: 3-Party Secure Framework for Customized Binary Neural Networks Inference](https://mlsanity.com/paper/2412.16449/) (2024)
- [DeepCQ: General-Purpose Deep-Surrogate Framework for Lossy Compression Quality Prediction](https://mlsanity.com/paper/2512.21433/) (2025)

## Abstract (arXiv)

Datasets for training object recognition systems are steadily increasing in size. This paper investigates the question of whether existing detectors will continue to improve as data grows, or saturate in performance due to limited model complexity and the Bayes risk associated with the feature spaces in which they operate. We focus on the popular paradigm of discriminatively trained templates defined on oriented gradient features. We investigate the performance of mixtures of templates as the number of mixture components and the amount of training data grows. Surprisingly, even with proper treatment of regularization and "outliers", the performance of classic mixture models appears to saturate quickly (~10 templates and ~100 positive training examples per template). This is not a limitation of the feature space as compositional mixtures that share template parameters via parts and that can synthesize new templates not encountered during training yield significantly better performance. Based on our analysis, we conjecture that the greatest gains in detection performance will continue to derive from improved representations and learning algorithms that can make efficient use of large datasets.