Yasin Bakis

M. Maruf, Arka Daw, Kazi Sajeed Mehrab et al. · microsoft-research

Images are increasingly becoming the currency for documenting biodiversity on the planet, providing novel opportunities for accelerating scientific discoveries in the field of organismal biology, especially with the advent of large vision-language models (VLMs). We ask if pre-trained VLMs can aid scientists in answering a range of biologically relevant questions without any additional fine-tuning. In this paper, we evaluate the effectiveness of 12 state-of-the-art (SOTA) VLMs in the field of organismal biology using a novel dataset, VLM4Bio, consisting of 469K question-answer pairs involving 30K images from three groups of organisms: fishes, birds, and butterflies, covering five biologically relevant tasks. We also explore the effects of applying prompting techniques and tests for reasoning hallucination on the performance of VLMs, shedding new light on the capabilities of current SOTA VLMs in answering biologically relevant questions using images. The code and datasets for running all the analyses reported in this paper can be found at https://github.com/sammarfy/VLM4Bio.

Kazi Sajeed Mehrab, M. Maruf, Arka Daw et al.

We introduce Fish-Visual Trait Analysis (Fish-Vista), the first organismal image dataset designed for the analysis of visual traits of aquatic species directly from images using problem formulations in computer vision. Fish-Vista contains 69,126 annotated images spanning 4,154 fish species, curated and organized to serve three downstream tasks of species classification, trait identification, and trait segmentation. Our work makes two key contributions. First, we perform a fully reproducible data processing pipeline to process images sourced from various museum collections. We annotate these images with carefully curated labels from biological databases and manual annotations to create an AI-ready dataset of visual traits, contributing to the advancement of AI in biodiversity science. Second, our proposed downstream tasks offer fertile grounds for novel computer vision research in addressing a variety of challenges such as long-tailed distributions, out-of-distribution generalization, learning with weak labels, explainable AI, and segmenting small objects. We benchmark the performance of several existing methods for our proposed tasks to expose future research opportunities in AI for biodiversity science problems involving visual traits.

Dom Jebbia, Xiaojun Wang, Yasin Bakis et al.

Flexible metadata pipelines are crucial for supporting the FAIR data principles. Despite this need, researchers seldom report their approaches for identifying metadata standards and protocols that support optimal flexibility. This paper reports on an initiative targeting the development of a flexible metadata pipeline for a collection containing over 300,000 digital fish specimen images, harvested from multiple data repositories and fish collections. The images and their associated metadata are being used for AI-related scientific research involving automated species identification, segmentation and trait extraction. The paper provides contextual background, followed by the presentation of a four-phased approach involving: 1. Assessment of the Problem, 2. Investigation of Solutions, 3. Implementation, and 4. Refinement. The work is part of the NSF Harnessing the Data Revolution, Biology Guided Neural Networks (NSF/HDR-BGNN) project and the HDR Imageomics Institute. An RDF graph prototype pipeline is presented, followed by a discussion of research implications and conclusion summarizing the results.