AGENTIC-SYSTEMSOctober 24, 20251 min read20 views

How AI is helping advance the science of bioacoustics to save endangered species

Google has released an updated version of Perch, an AI model designed to help conservationists analyze bioacoustic data more efficiently to protect endangered species and monitor ecosystem health.

Our new Perch model helps conservationists analyze audio faster to protect endangered species, from Hawaiian honeycreepers to coral reefs.

One of the ways scientists protect the health of our planet’s wild ecosystems is by using microphones (or underwater hydrophones) to collect vast amounts of audio dense with vocalizations from birds, frogs, insects, whales, fish and more. These recordings can tell us a lot about the animals present in a given area, along with other clues about the health of that ecosystem. Making sense of so much data, however, remains a massive undertaking.

Today, we are releasing an update to Perch, our AI model designed to help conservationists analyze bioacoustic data. This new model has better state-of-the-art off-the-shelf bird species predictions than the previous model. It can better adapt to new environments, particularly underwater ones like coral reefs. It’s trained on a wider range of animals, including mammals, amphibians and anthropogenic noise — nearly twice as much data in all, from public sources like Xeno-Canto and iNaturalist. It can disentangle complex acoustic scenes over thousands or even millions of hours of audio data. And it’s versatile, able to help answer many different kinds of questions, from “how many babies are being born” to “how many individual animals are present in a given area.”

In order to help scientists protect our planet’s ecosystems, we’re releasing this new version of Perch as an open model and making it available on Kaggle.

Source: Google DeepMind Blog

More in this category

agentic-systems

Generative Ontology Induction: Domain-Agnostic Schema Discovery from Document Corpora Using Large Language Models

Researchers introduce Generative Ontology Induction (GOI), a domain-agnostic framework that automatically extracts structured ontologies from document corpora using LLMs. Achieving 95-100% structural coverage, GOI addresses a major bottleneck in knowledge-intensive AI systems.

agentic-systems

JUMP: Single-Pass Membership Inference on Fine-Tuned Diffusion Language Models

Researchers have proposed JUMP, a novel single-pass membership inference attack designed for fine-tuned discrete diffusion language models (dLLMs). By leveraging the unique properties of dLLMs, JUMP significantly improves detection accuracy while drastically reducing the number of required queries.

agentic-systems

Masked Diffusion Language Models are Strong and Steerable Text-Based World Models for Agentic RL

Researchers demonstrate that Masked Diffusion Language Models (MDLMs) serve as highly effective, steerable text-based world models for agentic reinforcement learning. By leveraging bidirectional denoising, MDLMs outperform autoregressive models four times their size in coherence, groundedness, and rollout diversity.

NOW LET US Related – Democratizing AI with Small Language Models: Structured Benchmarking and Parameter-Efficient Fine-Tuning for Local Deployment

agentic-systems

Democratizing AI with Small Language Models: Structured Benchmarking and Parameter-Efficient Fine-Tuning for Local Deployment

A new study demonstrates that small language models (SLMs) under 3 billion parameters can serve as highly capable local experts for specialized tasks. By combining structured benchmarking with low-cost parameter-efficient fine-tuning (PEFT), institutions can achieve AI autonomy without relying on expensive hardware.

agentic-systems

PlanFlip: Attacking Multi-Agent LLM Systems via Planning-Phase Prompt Injection

Researchers have introduced PlanFlip, a novel prompt injection attack framework targeting the planning phase of multi-agent LLM systems. The study reveals critical security blind spots in homogeneous agent pipelines and demonstrates that reasoning-augmented models like DeepSeek-R1 exhibit strong resistance.

agentic-systems

Rater State Bias in RLHF Preference Data: An Audit Framework

A new study identifies a structured bias in Reinforcement Learning from Human Feedback (RLHF) caused by the psychological state of human raters. Under stress, raters' shifting preferences can propagate through reward modeling, potentially compromising AI policy optimization.