Legal

Open Source Credits

VeysaLabs is built on open source software and peer-reviewed research. This page acknowledges the projects and authors whose work makes the platform possible.

AI & Inference Saliency prediction and model infrastructure
TranSalNet
Transformer-based saliency prediction model. Core inference engine for VeysaLabs Creative Effectiveness Reports. Trained on the SALICON dataset.
Lou et al. (2022) · github.com/LJOVO/TranSalNet
Research Licence
SALICON Dataset
Large-scale saliency dataset used to train the TranSalNet model. Collected via crowdsourced mouse-tracking as a proxy for eye fixations.
Jiang et al. (2015) · salicon.net
Research Use
ML & Inference Packages Python packages running inside the GPU inference container
PyTorch
Core ML framework executing the TranSalNet model on GPU via Modal.
Meta AI / PyTorch Foundation · pytorch.org · torch==2.2.2
BSD 3-Clause
torchvision
PyTorch image transforms and preprocessing utilities used in the inference pipeline.
Meta AI / PyTorch Foundation · pytorch.org/vision
BSD 3-Clause
Pillow
Image loading, resizing, and preprocessing throughout the analysis pipeline.
Alex Clark and contributors · python-pillow.org
HPND
NumPy
Array operations and saliency map serialisation (numpy.save / numpy.load) in the inference pipeline.
NumPy contributors · numpy.org
BSD 3-Clause
SciPy
Scientific computing library used internally by TranSalNet during inference.
SciPy contributors · scipy.org
BSD 3-Clause
scikit-image
Image processing library used internally by TranSalNet during inference.
scikit-image contributors · scikit-image.org
BSD 3-Clause
Backend Server, API, and browser automation
FastAPI
Modern, high-performance Python web framework for building the VeysaLabs API.
Sebastián Ramírez · fastapi.tiangolo.com
MIT
Playwright
Headless browser automation used for URL capture — rendering and screenshotting web pages submitted for analysis.
Copyright Microsoft Corporation · playwright.dev
MIT
Python
Core backend language powering the inference pipeline, API server, and data processing.
Python Software Foundation · python.org
PSF Licence
Research Foundations Peer-reviewed science underpinning the analysis methodology
Pieters, R. & Wedel, M. (2004). Attention capture and transfer in advertising: Brand, pictorial, and text-size effects. Journal of Marketing, 68(2), 36–50.
Wedel, M. & Pieters, R. (2008). A review of eye-tracking research in marketing. Review of Marketing Research, 4, 123–147.
Xie, S. et al. (2023). Predicting visual attention in graphic design: A computational approach. Full citation to be confirmed before publishing.
Jiang, M. et al. (2015). SALICON: Saliency in context. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 1072–1080.
Lou, P. et al. (2022). TranSalNet: Towards perceptually relevant visual saliency prediction. Neurocomputing, 494, 455–467.