arXiv subjects

Zi K. Deng

Publications and source records attributed to Zi K. Deng.

3 recordsLinked to original sources

arXiv2026

WeedNet: A Foundation Model-Based Global-to-Local AI Approach for Real-Time Weed Species Identification and Classification

Early weed identification is crucial for effective management and control, and researchers, agronomists, and technology developers are increasingly interested in automating this process using computer vision and artificial intelligence; however, limited expert-verified data and variable morphological features have hindered the development of AI-based weed identification models. To address these issues, we present WeedNet, a global-scale weed identification model that can recognize an extensive set of weed species. WeedNet is an end-to-end real-time weed identification pipeline that uses self-supervised learning, fine-tuning, and enhanced trustworthiness strategies. WeedNet achieved 91.02% accuracy across 1,593 weed species, with 41% of species achieving 100% accuracy. Using a fine-tuning approach, the local Iowa WeedNet model achieved 97.38% overall accuracy for 84 Iowa weeds. Testing across intra-species dissimilarity and inter-species similarity suggests that diversity in the collected image, spanning all growth stages and distinct plant characteristics, is crucial to driving model performance. The global WeedNet model serves as a foundation, and the global-to-local strategy enables targeted fine-tuning to improve performance in regional weed communities. Additional validation of drone- and ground-rover-based images highlights WeedNet's potential for integration into robotic platforms. Furthermore, integration with artificial intelligence for conversational use provides intelligent agricultural and ecological conservation consulting tools for farmers, researchers, and government agencies across diverse landscapes.

cs.CV↗

arXiv2025

TerraIncognita: A Dynamic Benchmark for Species Discovery Using Frontier Models

The rapid global loss of biodiversity, particularly among insects, represents an urgent ecological crisis. Current methods for insect species discovery are manual, slow, and severely constrained by taxonomic expertise, hindering timely conservation actions. We introduce TerraIncognita, a dynamic benchmark designed to evaluate state-of-the-art multimodal models for the challenging problem of identifying unknown, potentially undescribed insect species from image data. Our benchmark dataset combines a mix of expertly annotated images of insect species likely known to frontier AI models, and images of rare and poorly known species, for which few/no publicly available images exist. These images were collected from underexplored biodiversity hotspots, realistically mimicking open-world discovery scenarios faced by ecologists. The benchmark assesses models' proficiency in hierarchical taxonomic classification, their capability to detect and abstain from out-of-distribution (OOD) samples representing novel species, and their ability to generate explanations aligned with expert taxonomic knowledge. Notably, top-performing models achieve over 90\% F1 at the Order level on known species, but drop below 2\% at the Species level, highlighting the sharp difficulty gradient from coarse to fine taxonomic prediction (Order $\rightarrow$ Family $\rightarrow$ Genus $\rightarrow$ Species). TerraIncognita will be updated regularly, and by committing to quarterly dataset expansions (of both known and novel species), will provide an evolving platform for longitudinal benchmarking of frontier AI methods. All TerraIncognita data, results, and future updates are available \href{https://baskargroup.github.io/TerraIncognita/}{here}.

cs.CV↗

arXiv2025

BioTrove: A Large Curated Image Dataset Enabling AI for Biodiversity

We introduce BioTrove, the largest publicly accessible dataset designed to advance AI applications in biodiversity. Curated from the iNaturalist platform and vetted to include only research-grade data, BioTrove contains 161.9 million images, offering unprecedented scale and diversity from three primary kingdoms: Animalia ("animals"), Fungi ("fungi"), and Plantae ("plants"), spanning approximately 366.6K species. Each image is annotated with scientific names, taxonomic hierarchies, and common names, providing rich metadata to support accurate AI model development across diverse species and ecosystems. We demonstrate the value of BioTrove by releasing a suite of CLIP models trained using a subset of 40 million captioned images, known as BioTrove-Train. This subset focuses on seven categories within the dataset that are underrepresented in standard image recognition models, selected for their critical role in biodiversity and agriculture: Aves ("birds"), Arachnida ("spiders/ticks/mites"), Insecta ("insects"), Plantae ("plants"), Fungi ("fungi"), Mollusca ("snails"), and Reptilia ("snakes/lizards"). To support rigorous assessment, we introduce several new benchmarks and report model accuracy for zero-shot learning across life stages, rare species, confounding species, and multiple taxonomic levels. We anticipate that BioTrove will spur the development of AI models capable of supporting digital tools for pest control, crop monitoring, biodiversity assessment, and environmental conservation. These advancements are crucial for ensuring food security, preserving ecosystems, and mitigating the impacts of climate change. BioTrove is publicly available, easily accessible, and ready for immediate use.

cs.CV↗