# PHOEBI: An Open-World Benchmark for Multi-Label Bacterial Identification in Phase-Contrast Microscopy

arXiv: https://arxiv.org/abs/2606.22890 · HTML: https://mlsanity.com/paper/2606.22890/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 38 references that are in the map.
- First posted on arXiv 22 Jun 2026.

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

- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [mixup: Beyond Empirical Risk Minimization](https://mlsanity.com/paper/1710.09412/) (2017)
- [Prototypical Networks for Few-shot Learning](https://mlsanity.com/paper/1703.05175/) (2017)
- [A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks](https://mlsanity.com/paper/1610.02136/) (2016)
- [Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distances](https://mlsanity.com/paper/1306.0895/) (2013)
- [Datasheets for Datasets](https://mlsanity.com/paper/1803.09010/) (2018)
- [Energy-based Out-of-distribution Detection](https://mlsanity.com/paper/2010.03759/) (2020)
- [Attention-based Deep Multiple Instance Learning](https://mlsanity.com/paper/1802.04712/) (2018)
- [Deep Anomaly Detection with Outlier Exposure](https://mlsanity.com/paper/1812.04606/) (2018)
- [Out-of-Distribution Detection with Deep Nearest Neighbors](https://mlsanity.com/paper/2204.06507/) (2022)
- [BiomedCLIP: a multimodal biomedical foundation model pretrained from fifteen million scientific image-text pairs](https://mlsanity.com/paper/2303.00915/) (2023)
- [A Unifying Review of Deep and Shallow Anomaly Detection](https://mlsanity.com/paper/2009.11732/) (2020)
- [Asymmetric Loss For Multi-Label Classification](https://mlsanity.com/paper/2009.14119/) (2020)
- [From Softmax to Sparsemax: A Sparse Model of Attention and Multi-Label Classification](https://mlsanity.com/paper/1602.02068/) (2016)
- [Learning to Discover Novel Visual Categories via Deep Transfer Clustering](https://mlsanity.com/paper/1908.09884/) (2019)
- [Generalized Category Discovery](https://mlsanity.com/paper/2201.02609/) (2022)
- [CNN-RNN: A Unified Framework for Multi-label Image Classification](https://mlsanity.com/paper/1604.04573/) (2016)
- [A Unified Objective for Novel Class Discovery](https://mlsanity.com/paper/2108.08536/) (2021)
- [Query2Label: A Simple Transformer Way to Multi-Label Classification](https://mlsanity.com/paper/2107.10834/) (2021)
- [Parametric Classification for Generalized Category Discovery: A Baseline Study](https://mlsanity.com/paper/2211.11727/) (2022)
- [Modeling Inter-Class and Intra-Class Constraints in Novel Class Discovery](https://mlsanity.com/paper/2210.03591/) (2022)
- [Class-relation Knowledge Distillation for Novel Class Discovery](https://mlsanity.com/paper/2307.09158/) (2023)
- [μ-Bench: A Vision-Language Benchmark for Microscopy Understanding](https://mlsanity.com/paper/2407.01791/) (2024)
- [Novel Class Discovery for Ultra-Fine-Grained Visual Categorization](https://mlsanity.com/paper/2405.06283/) (2024)
- [Overview of PlantCLEF 2024: multi-species plant identification in vegetation plot images](https://mlsanity.com/paper/2509.15768/) (2025)

## Abstract (arXiv)

Optical microscopy (OM) enables rapid, label-free imaging of live bacteria and is the standard instrument for species identification across clinical, environmental, and industrial microbiology. Real samples, however, are routinely polymicrobial and may contain organisms never seen during training, and no computer-vision benchmark evaluates multi-label species identification from phase-contrast microscopy (PCM) of such mixtures. We introduce Phase-contrast Optical bEnchmark for Bacterial Identification (PHOEBI), a wet-lab-prepared dataset of 120,000 PCM images covering 40 combinations of six rod-shaped species, together with a leave-combinations-out (LCO) protocol that holds out entire species combinations, mirroring a model trained on catalogued mixtures that must recognise new ones. Under LCO, gradient-trained per-image classifiers, from fine-tuned backbones to attention-based multiple-instance learning, collapse on unseen combinations despite high in-distribution accuracy, and the failure lies in how per-image predictions are aggregated rather than in the visual representation. We propose three lightweight anchor-based decoders that read each species' presence against fixed geometric prototypes over a shared frozen tile-feature pool, and they remain stable under the same shift. Without additional training, the same features also support open-set rejection of unseen species and the discovery of a new class from unlabeled test images, with negligible disruption to the known classes.