Projects
Diffusion Models for Flexible Packaging
Our current research initiative exploring generative models for industrial design applications.
Under this umbrella, we are working on three sub-projects addressing physical and perceptual constraints in print-ready image synthesis.
RobustSep
PPP-conditioned RGB to CMYK separation with constraint refinement
Separation is conditioned on a Process-and-Press Prior (PPP) and refined through explicit constraints rather than treated as unconstrained translation.
Conservative ink-limit logic (TAC and per-channel caps), overprint sanity, neutral/dark stability, separation regularity.
Patch-aware (16×16), proposes multiple candidates per patch, applies robustness checks under parametric perturbations (tone curves, ink strength). Outputs CMYK result and concise decision report (risk flags and constraint tightness).
TAC violation rate, per-channel cap adherence, overprint stability, neutral preservation (ΔE), robustness score under PPP perturbations.
Retained for private usage. Will be published after redaction.
Deep Reinforcement Learning Halftoning
Halftoning under print constraints via deep reinforcement learning
Halftoning framed as constrained optimization with deep reinforcement learning as comparator to classical baselines.
Per-channel frequency bounds, dot gain limits, moiré avoidance, local tone linearity.
JJN error diffusion, void-and-cluster blue-noise ordered dithering, Direct Binary Search.
Constraint pass rate, robustness under gain variation, perceptual similarity (SSIM, LPIPS), per-channel frequency analysis.
Retained for private usage. Will be published after redaction.
Inference-Time Loss-Guided Colour Preservation in Diffusion Sampling
Inference-time constraint enforcement for perceptual colour fidelity
Maintaining perceptual colour fidelity when generative processes introduce stochastic variations that can drift from intended colour specifications.
Perceptual colour drift bounds (ΔE metrics), region-specific preservation, semantic colour consistency across generated outputs.
Inference-time losses and regularizers with explicit monitoring of colour drift and perceptual reconstruction error.
Perceptual colour drift (ΔE), region preservation metrics, semantic consistency scores, constraint violation frequency.
@article{colourperception2025,
title={Inference-Time Loss-Guided Colour Preservation in Diffusion Sampling},
author={Ahuja, Angad Singh and Patel, Poorva and Anandh, Aarush Ram},
journal={arXiv preprint},
year={2025},
note={Pre-print, out now}
}