# Reconstructed holograms and explanation-aware evaluation for low-cost computational pollen analysis in veterinary cytology

arXiv: https://arxiv.org/abs/2609.23017 · HTML: https://mlsanity.com/paper/2609.23017/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 56 references that are in the map.
- First posted on arXiv 19 Sep 2026.

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

- [Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization](https://mlsanity.com/paper/1610.02391/) (2016)
- [Axiomatic Attribution for Deep Networks](https://mlsanity.com/paper/1703.01365/) (2017)
- [Learning Deep Features for Discriminative Localization](https://mlsanity.com/paper/1512.04150/) (2015)
- [Visualizing and Understanding Convolutional Networks](https://mlsanity.com/paper/1311.2901/) (2013)
- [Spectral Normalization for Generative Adversarial Networks](https://mlsanity.com/paper/1802.05957/) (2018)
- [Sanity Checks for Saliency Maps](https://mlsanity.com/paper/1810.03292/) (2018)
- [SmoothGrad: removing noise by adding noise](https://mlsanity.com/paper/1706.03825/) (2017)
- [RISE: Randomized Input Sampling for Explanation of Black-box Models](https://mlsanity.com/paper/1806.07421/) (2018)
- [Score-CAM: Score-Weighted Visual Explanations for Convolutional Neural Networks](https://mlsanity.com/paper/1910.01279/) (2019)
- [Not Just a Black Box: Learning Important Features Through Propagating Activation Differences](https://mlsanity.com/paper/1605.01713/) (2016)
- [Explainable artificial intelligence (XAI) in deep learning-based medical image analysis](https://mlsanity.com/paper/2107.10912/) (2021)
- [MobileNetV4 -- Universal Models for the Mobile Ecosystem](https://mlsanity.com/paper/2404.10518/) (2024)
- [Black-box Explanation of Object Detectors via Saliency Maps](https://mlsanity.com/paper/2006.03204/) (2020)
- [Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models](https://mlsanity.com/paper/2606.03748/) (2026)
- [Phase recovery and holographic image reconstruction using deep learning in neural networks](https://mlsanity.com/paper/1705.04286/) (2017)
- [Group-CAM: Group Score-Weighted Visual Explanations for Deep Convolutional Networks](https://mlsanity.com/paper/2103.13859/) (2021)
- [Towards Better Explanations for Object Detection](https://mlsanity.com/paper/2306.02744/) (2023)
- [MFPP: Morphological Fragmental Perturbation Pyramid for Black-Box Model Explanations](https://mlsanity.com/paper/2006.02659/) (2020)
- [G-CAME: Gaussian-Class Activation Mapping Explainer for Object Detectors](https://mlsanity.com/paper/2306.03400/) (2023)
- [Explaining YOLO: Leveraging Grad-CAM to Explain Object Detections](https://mlsanity.com/paper/2211.12108/) (2022)
- [Model-agnostic explainable artificial intelligence for object detection in image data](https://mlsanity.com/paper/2303.17249/) (2023)
- [Physics-driven universal twin-image removal network for digital in-line holographic microscopy](https://mlsanity.com/paper/2308.04471/) (2023)
- [On the Black-box Explainability of Object Detection Models for Safe and Trustworthy Industrial Applications](https://mlsanity.com/paper/2411.00818/) (2024)
- [Automated Pollen Recognition in Optical and Holographic Microscopy Images](https://mlsanity.com/paper/2512.08589/) (2025)
- [AI-Augmented Pollen Recognition in Optical and Holographic Microscopy for Veterinary Imaging](https://mlsanity.com/paper/2512.12101/) (2025)

## Abstract (arXiv)

Automated pollen analysis supports veterinary cytology, but brightfield microscopy is costlier and more complex than lens-less digital in-line holographic microscopy. We evaluate whether reconstructed holograms can narrow this gap and whether model explanations remain reliable under modality change. Six pollen species were imaged by brightfield and holographic microscopy. Raw, single back-propagation and iterative phase retrieval holograms were evaluated with YOLOv26s detection and MobileNetV4 classification after anchor-based annotation transfer. Six attribution methods were assessed for spatial grounding and faithfulness with the Attribution Health Inspection and Repair (AHIR) protocol, which tests model brittleness under weak noise and corrects attribution-map granularity when needed. Brightfield achieved 0.6890 mAP50-95 (0.8865 mAP50) for detection and 0.9687 macro-F1 (0.9705 accuracy) for classification. Reconstructed holograms narrowed the gap with a task-dependent split: p-type was strongest for detection at 0.5324 mAP50-95 (0.8229 mAP50), while r-type was strongest for classification at 0.7695 macro-F1 (0.7866 accuracy), both far above raw-hologram baselines. Activation-based explanations localized strongly on grains, and region-based methods retained ~60 to ~80% of faithfulness under holography. The holographic detector was highly brittle to weak perturbations, saturating deletion-based evaluation while insertion remained informative. Pixel-level gradient explanations approached random floor, yet spatial smoothing restored p-type gradient faithfulness from 0.05 to 0.51. For holographic classification, perturbation-based explanations remained faithful while gradient-based methods fell below random floor. Reconstruction improves low-cost holographic pollen analysis, while AHIR distinguishes genuine attribution failure from artifacts caused by model brittleness and map granularity.