The field is witnessing a significant shift towards developing more efficient and accurate methods for statistical estimation and computational imaging. Researchers are exploring new approaches to improve the convergence rates of various algorithms, including flow matching and diffusion models. The use of novel techniques, such as normalizing flows and convex-relaxation-based algorithms, is becoming increasingly popular. These methods have been shown to achieve state-of-the-art results in estimating smooth distributions and reconstructing molecular structures from noisy tomographic projection images. Noteworthy papers include: On Flow Matching KL Divergence, which derives a deterministic upper bound on the KL divergence of flow-matching distribution approximation. Two Datasets Are Better Than One: Method of Double Moments for 3-D Reconstruction in Cryo-EM, which introduces a new data fusion framework for reconstructing molecular structures from multiple datasets. FMMI: Flow Matching Mutual Information Estimation, which presents a novel mutual information estimator that reframes the discriminative approach.