# Heuristic Style Transfer for Real-Time, Efficient Weather Attribute Detection

arXiv: https://arxiv.org/abs/2604.13947 · HTML: https://mlsanity.com/paper/2604.13947/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 60 references that are in the map.
- First posted on arXiv 15 Apr 2026.

## Builds on (most cited first, 25 of 25)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [Deep Unsupervised Learning using Nonequilibrium Thermodynamics](https://mlsanity.com/paper/1503.03585/) (2015)
- [Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization](https://mlsanity.com/paper/1610.02391/) (2016)
- [Generative Adversarial Networks](https://mlsanity.com/paper/2203.00667/) (2022)
- [Squeeze-and-Excitation Networks](https://mlsanity.com/paper/1709.01507/) (2017)
- [Image-to-Image Translation with Conditional Adversarial Networks](https://mlsanity.com/paper/1611.07004/) (2016)
- [Non-local Neural Networks](https://mlsanity.com/paper/1711.07971/) (2017)
- [An Empirical Study of Training Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.02057/) (2021)
- [Show, Attend and Tell: Neural Image Caption Generation with Visual Attention](https://mlsanity.com/paper/1502.03044/) (2015)
- [A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on Reasoning, Hallucination, and Interactivity](https://mlsanity.com/paper/2302.04023/) (2023)
- [Contrastive Learning for Unpaired Image-to-Image Translation](https://mlsanity.com/paper/2007.15651/) (2020)
- [Theoretical Foundations of t-SNE for Visualizing High-Dimensional Clustered Data](https://mlsanity.com/paper/2105.07536/) (2021)
- [Weather and Light Level Classification for Autonomous Driving: Dataset, Baseline and Active Learning](https://mlsanity.com/paper/2104.14042/) (2021)
- [WeatherNet: Recognising weather and visual conditions from street-level images using deep residual learning](https://mlsanity.com/paper/1910.09910/) (2019)
- [ChatGPT: Vision and Challenges](https://mlsanity.com/paper/2305.15323/) (2023)
- [Weather Classification: A new multi-class dataset, data augmentation approach and comprehensive evaluations of Convolutional Neural Networks](https://mlsanity.com/paper/1808.00588/) (2018)
- [DMRVisNet: Deep Multi-head Regression Network for Pixel-wise Visibility Estimation Under Foggy Weather](https://mlsanity.com/paper/2112.04278/) (2021)
- [Real-Time Environment Condition Classification for Autonomous Vehicles](https://mlsanity.com/paper/2405.19305/) (2024)
- [Beyond Text: Unveiling Multimodal Proficiency of Large Language Models with MultiAPI Benchmark](https://mlsanity.com/paper/2311.13053/) (2023)
- [Low-latency Perception in Off-Road Dynamical Low Visibility Environments](https://mlsanity.com/paper/2012.13014/) (2020)
- [SemanticSpray++: A Multimodal Dataset for Autonomous Driving in Wet Surface Conditions](https://mlsanity.com/paper/2406.09945/) (2024)
- [Robust ADAS: Enhancing Robustness of Machine Learning-based Advanced Driver Assistance Systems for Adverse Weather](https://mlsanity.com/paper/2407.02581/) (2024)
- [Real-Time Weather Image Classification with SVM](https://mlsanity.com/paper/2409.00821/) (2024)

## Abstract (arXiv)

We present lightweight and efficient architectures to detect weather conditions from RGB images, predicting the weather type (sunny, rain, snow, fog) and 11 complementary attributes such as intensity, visibility, and ground condition, for a total of 53 classes across the tasks. This work examines to what extent weather conditions manifest as variations in visual style. We investigate style-inspired techniques, including Gram matrices, a truncated ResNet-50 targeting lower and intermediate layers, and PatchGAN-style architectures, within a multi-task framework with attention mechanisms. Two families are introduced: RTM (ResNet50-Truncated-MultiTasks) and PMG (PatchGAN-MultiTasks-Gram), together with their variants. Our contributions include automation of Gram-matrix computation, integration of PatchGAN into supervised multi-task learning, and local style capture through local Gram for improved spatial coherence. We also release a dataset of 503,875 images annotated with 12 weather attributes under a Creative Commons Attribution (CC-BY) license. The models achieve F1 scores above 96 percent on our internal test set and above 78 percent in zero-shot evaluation on several external datasets, confirming their generalization ability. The PMG architecture, with fewer than 5 million parameters, runs in real time with a small memory footprint, making it suitable for embedded systems. The modular design of the models also allows style-related or weather-related tasks to be added or removed as needed.