This is SPIRAL-TAP: Sparse Poisson Intensity Reconstruction ALgorithms - Theory and Practice

The observations in many applications consist of counts of discrete events, such as photons hitting a detector, which cannot be effectively modeled using an additive bounded or Gaussian noise model, and instead require a Poisson noise model. As a result, accurate reconstruction of a spatially or temporally distributed phenomenon (f*) from Poisson data (y) cannot be effectively accomplished by minimizing a conventional penalized least-squares objective function. The problem addressed in this paper is the estimation of f* from y in an inverse problem setting, where (a) the number of unknowns may potentially be larger than the number of observations and (b) f* admits a sparse approximation. The optimization formulation considered in this paper uses a penalized negative Poisson log-likelihood objective function with nonnegativity constraints (since Poisson intensities are naturally nonnegative). In particular, the proposed approach incorporates key ideas of using separable quadratic approximations to the objective function at each iteration and penalization terms related to l1 norms of coefficient vectors, total variation seminorms, and partition-based multiscale estimation methods.

💡 Research Summary

The paper introduces SPIRAL‑TAP (Sparse Poisson Intensity Reconstruction ALgorithms – Theory and Practice), a unified optimization framework for recovering a non‑negative intensity field f* from Poisson‑distributed measurements y in ill‑posed inverse problems where the number of unknowns may exceed the number of observations and the underlying signal admits a sparse representation. The authors argue that conventional penalized least‑squares approaches, which assume additive Gaussian noise, are fundamentally mismatched to count data because Poisson noise is signal‑dependent, asymmetric, and discrete. Consequently, they formulate the reconstruction problem as the minimization of a penalized negative Poisson log‑likelihood

  min_{f≥0} ℓ(f) + R(f)

where

  ℓ(f)=∑_{k=1}^{m}(