Menu
Sign In Search Podcasts Charts People & Topics Add Podcast API Pricing
Podcast Image

AI可可AI生活

[人人能懂] 从看见空间、探索信息到理解“不要”

18 Nov 2025

Description

你有没有想过,能写诗作画的AI,为什么有时却像个固执的孩子?本期我们要聊的几篇最新论文,就试图教会AI一些我们习以为常、但它却难以理解的人类智慧。我们将一起看看,如何治好AI的“路痴”症,让它拥有空间感;如何让它从被动看图,变身主动破案的“侦探”;甚至,如何通过巧妙的“换个姿势”,让它终于听懂“不要”,并随心所欲地调整观察事物的“粒度”。00:00:33 人工智能的“路痴”难题00:05:24 AI侦探,如何给千米大桥做“体检”?00:09:59 从“你猜”到“你定”:AI图像分割的新玩法00:14:45 换个姿势,让AI听懂“不要”本期介绍的几篇论文:[CV] Scaling Spatial Intelligence with Multimodal Foundation Models  [SenseTime Research]  https://arxiv.org/abs/2511.13719 ---[CV] BridgeEQA: Virtual Embodied Agents for Real Bridge Inspections  [University of Houston]  https://arxiv.org/abs/2511.12676 ---[CV] UnSAMv2: Self-Supervised Learning Enables Segment Anything at Any Granularity  [UC Berkeley]  https://arxiv.org/abs/2511.13714 ---[CV] SpaceVLM: Sub-Space Modeling of Negation in Vision-Language Models  [MIT]  https://arxiv.org/abs/2511.12331 

Audio
Featured in this Episode

No persons identified in this episode.

Transcription

This episode hasn't been transcribed yet

Help us prioritize this episode for transcription by upvoting it.

0 upvotes
🗳️ Sign in to Upvote

Popular episodes get transcribed faster

Comments

There are no comments yet.

Please log in to write the first comment.