#多模态

共 39 篇带此标签

39 articles Newest 2026-11-21
Archive · #多模态 11 articles
Tools

dettato’s 21M Audio-to-MIDI Model Beats 1.3B

dettato claims strong audio-to-MIDI results with roughly 21 million parameters and a model size about 60 times smaller than muscriptor 1.3b. Its open weights make the project easy to test, though independent benchmarks, training details, and licensing remain necessary to verify.

2026-11-21 Source 2026-11-17 7 min 36 views
Products

Genie 4 and the Gap to a Real World Model

Google DeepMind has released Genie 4 through Project Genie, enabling interactive environments generated from text, images, or video. Before it can serve as a dependable training ground for AI agents, it still has to overcome gaps in physics, long-horizon consistency, and fidelity to the real world.

2026-11-20 Source 2026-11-18 7 min 36 views
Case Studies

OpenAI Disbands Robotics Team as General Models Take Over

OpenAI has disbanded its robotics team as general-purpose models begin operating robots, according to Fortune. The restructuring signals a possible shift in strategy, but it does not establish that OpenAI is leaving the robotics market or that the new approach has proven viable.

2026-11-14 Source 2026-11-13 8 min 36 views
Case Studies

48 AI Agents Searched for a Maya Pyramid in 72 Hours

A Google SWE system used 48 AI Agents to screen NASA lidar data during a 72-hour archaeological search. The engineering challenge was not the discovery itself, but how the Agents divided the work, validated evidence, and amplified errors.

2026-11-12 Source 2026-11-11 7 min 36 views
Case Studies

How an AI Agent Helped Discover a Maya Site

Pietro Miotti’s Agent team helped locate a previously unmapped Maya site. The more revealing story is how the system connected online research, field investigation, and evidence preservation while human archaeologists retained authority over consequential decisions.

2026-11-12 Source 2026-11-11 8 min 36 views
Case Studies

What $20,000 Really Buys for an Open-Weight Image Model

A two-person lab trained the PJIT-21K text-to-image model from scratch for roughly $20,000 and released its weights under Apache 2.0. This analysis breaks down training, inference, licensing, and deployment costs for teams evaluating open models.

2026-11-06 Source 2026-11-04 7 min 36 views
Opinion

LLM Voice: The Hard Parts of Real-Time Voice Agents

Martin Fowler examines how to build LLM-powered, real-time voice applications. The core challenge is not combining ASR, LLM, and TTS, but controlling end-to-end latency, interruptions, and system state.

2026-10-30 Source 2026-10-29 9 min 36 views
Tools

Hurston-1 and the Data Behind Regional Accents

Destined AI has released Hurston-1, a narrow-bandwidth TTS model focused on Southern American and Regional African English. Its licensed dataset and selective training slice reveal both the model's potential and the limits of current accent evaluation.

2026-10-26 Source 2026-10-25 7 min 36 views
Case Studies

CNN vs. GPT-4o and Claude 3.5 in Skin Cancer

In a Lynch Lab evaluation, a CNN developed by a 10-year-old researcher outperformed GPT-4o and Claude 3.5 on skin lesion classification. The finding raises broader questions about model size, specialized architecture, and the rigor of medical AI benchmarks.

2026-10-26 Source 2026-10-25 6 min 36 views