New article! Where are GIScience faculty hired from? Yanbing Chen et al. w/data @GISphereGuide analyze faculty mobility and research themes through hiring networks #GISchat doi.org/10.1080/15230406.202… Below where current GIS faculty are (x-axis) & where they got their PhDs (y-axis)

🎉 Two years after my graduation from Berkeley, my thesis is now published in Landscape and Urban Planning! We used LiDAR + microclimate simulation to reveal urban forest equity gaps in Portland, OR. 📄 Read: lnkd.in/gegZwn4W #UrbanForests #LiDAR #GIS #UrbanEquity

🌍🎧 𝗦𝗼𝘂𝗻𝗗𝗶𝗧: 𝗧𝘂𝗿𝗻𝗶𝗻𝗴 𝗦𝗼𝘂𝗻𝗱𝘀𝗰𝗮𝗽𝘀 𝗶𝗻𝘁𝗼 𝗥𝗲𝗮𝗹𝗶𝘀𝘁𝗶𝗰 𝗟𝗮𝗻𝗱𝘀𝗰𝗮𝗽𝗲𝘀! 🌿🏙️ Imagine synthesizing realistic geographic landscapes directly from soundscapes—bird chirping leading to lush parks, or bustling city noises bringing urban streets to life. Our team is thrilled to introduce 𝗚𝗲𝗼-𝗖𝗼𝗻𝘁𝗲𝘅𝘁𝘂𝗮𝗹 𝗦𝗼𝘂𝗻𝗱𝘀𝗰𝗮𝗽𝗲-𝘁𝗼-𝗟𝗮𝗻𝗱𝘀𝗰𝗮𝗽𝗲 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝗼𝗻 (GeoS2L), powered by our innovative SounDiT model. 🔊 𝗪𝗵𝘆 𝗚𝗲𝗼𝗦𝟮𝗟? Previous audio-to-image models often overlooked geographic contexts, producing mismatched or unrealistic images. SounDiT leverages geographic knowledge to create images that truly align with real-world settings. 🚀 𝗞𝗲𝘆 𝗜𝗻𝗻𝗼𝘃𝗮𝘁𝗶𝗼𝗻𝘀: • 𝗦𝗼𝘂𝗻𝗗𝗶𝗧 𝗠𝗼𝗱𝗲𝗹: A novel Diffusion Transformer integrating sound, scene prompts, and geographic data for unprecedented realism. • 𝗚𝗲𝗼-𝗰𝗼𝗻𝘁𝗲𝘅𝘁𝘂𝗮𝗹 𝗗𝗮𝘁𝗮𝘀𝗲𝘁𝘀: Created two large-scale datasets—SoundingSVI (330K pairs) and SonicUrban (250K pairs)—to support geographically coherent synthesis. • 𝗣𝗹𝗮𝗰𝗲 𝗦𝗶𝗺𝗶𝗹𝗮𝗿𝗶𝘁𝘆 𝗦𝗰𝗼𝗿𝗲 (𝗣𝗦𝗦): A multi-level evaluation framework ensuring generated images match soundscapes in element, scene, and human perception. 📈 𝗢𝘂𝘁𝘀𝘁𝗮𝗻𝗱𝗶𝗻𝗴 𝗥𝗲𝘀𝘂𝗹𝘁𝘀: • SounDiT significantly outperforms existing models in visual fidelity and geographic coherence. • Achieved remarkable improvements in landscape generation across multiple geo-contextual benchmarks. 🌐 𝗘𝘅𝗽𝗹𝗼𝗿𝗲 𝗼𝘂𝗿 𝗽𝗿𝗼𝗷𝗲𝗰𝘁: arxiv.org/abs/2505.12734 We're excited to explore more applications at the intersection of generative AI, geography, urban planning, and environmental sciences. Let's connect sound and vision together! @YuhaoKang

🚀 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗔𝗜 𝗶𝘀 𝘁𝗵𝗲 𝗡𝗲𝘅𝘁 𝗤𝘂𝗮𝗻𝘁𝘂𝗺 𝗟𝗲𝗮𝗽 𝗳𝗼𝗿 𝗔𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀 𝗗𝗿𝗶𝘃𝗶𝗻𝗴—𝗛𝗲𝗿𝗲’𝘀 𝘁𝗵𝗲 𝗛𝗶𝘁𝗰𝗵𝗵𝗶𝗸𝗲𝗿'𝘀 𝗚𝘂𝗶𝗱𝗲 𝗬𝗼𝘂’𝘃𝗲 𝗕𝗲𝗲𝗻 𝗪𝗮𝗶𝘁𝗶𝗻𝗴 𝗙𝗼𝗿! 🚀 We're thrilled to share our most comprehensive, 128-page survey, "𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗔𝗜 𝗳𝗼𝗿 𝗔𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀 𝗗𝗿𝗶𝘃𝗶𝗻𝗴: 𝗙𝗿𝗼𝗻𝘁𝗶𝗲𝗿𝘀 𝗮𝗻𝗱 𝗢𝗽𝗽𝗼𝗿𝘁𝘂𝗻𝗶𝘁𝗶𝗲𝘀," developed collaboratively by 23 institutions, including universities like #TAMU, #UCRiversie, #Purdue, #UMich, #UFL, hashtag#ASU, #Columbia, #Stanford, etc., and industry leaders such as #NVIDIA, #Bosch, #TTI, #Adobe, to name a few. 🔥 This isn't merely another overview; it’s an in-depth exploration that synthesizes breakthroughs, analyzes state-of-the-art technologies, and charts a clear, detailed roadmap for the future of autonomous driving and intelligent mobility. 🧩 𝗛𝗲𝗿𝗲'𝘀 𝗮 𝗱𝗲𝗲𝗽𝗲𝗿 𝗹𝗼𝗼𝗸 𝗶𝗻𝘁𝗼 𝘄𝗵𝗮𝘁 𝘄𝗲’𝘃𝗲 𝗰𝗼𝘃𝗲𝗿𝗲𝗱: 🔹 𝗙𝘂𝗻𝗱𝗮𝗺𝗲𝗻𝘁𝗮𝗹𝘀 𝗼𝗳 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗔𝗜: Clear overview of VAEs, GANs, Diffusion Models, NeRF, and LLMs. 🔹 𝗖𝘂𝘁𝘁𝗶𝗻𝗴-𝗘𝗱𝗴𝗲 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗔𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻𝘀 𝗶𝗻 𝗔𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀 𝗗𝗿𝗶𝘃𝗶𝗻𝗴: 𝘚𝘦𝘯𝘴𝘰𝘳 𝘋𝘢𝘵𝘢 𝘎𝘦𝘯𝘦𝘳𝘢𝘵𝘪𝘰𝘯; 𝘛𝘳𝘢𝘫𝘦𝘤𝘵𝘰𝘳𝘺 𝘢𝘯𝘥 𝘔𝘰𝘵𝘪𝘰𝘯 𝘗𝘳𝘦𝘥𝘪𝘤𝘵𝘪𝘰𝘯; 𝘖𝘤𝘤𝘶𝘱𝘢𝘯𝘤𝘺 𝘢𝘯𝘥 𝘚𝘤𝘦𝘯𝘦 𝘜𝘯𝘥𝘦𝘳𝘴𝘵𝘢𝘯𝘥𝘪𝘯𝘨; 𝘚𝘤𝘦𝘯𝘦 𝘌𝘥𝘪𝘵𝘪𝘯𝘨 𝘢𝘯𝘥 𝘊𝘶𝘴𝘵𝘰𝘮𝘪𝘻𝘢𝘵𝘪𝘰𝘯. 🔹 𝗥𝗲𝗮𝗹-𝗪𝗼𝗿𝗹𝗱 𝗔𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻𝘀 𝗮𝗻𝗱 𝗣𝗿𝗮𝗰𝘁𝗶𝗰𝗮𝗹 𝗜𝗻𝘁𝗲𝗴𝗿𝗮𝘁𝗶𝗼𝗻𝘀: 𝘌𝘯𝘥-𝘵𝘰-𝘌𝘯𝘥 𝘈𝘶𝘵𝘰𝘯𝘰𝘮𝘰𝘶𝘴 𝘚𝘺𝘴𝘵𝘦𝘮𝘴; 𝘚𝘺𝘯𝘵𝘩𝘦𝘵𝘪𝘤 𝘋𝘢𝘵𝘢 𝘞𝘰𝘳𝘬𝘧𝘭𝘰𝘸; and 𝘋𝘪𝘨𝘪𝘵𝘢𝘭 𝘛𝘸𝘪𝘯𝘴 𝘢𝘯𝘥 𝘚𝘮𝘢𝘳𝘵 𝘛𝘳𝘢𝘯𝘴𝘱𝘰𝘳𝘵𝘢𝘵𝘪𝘰𝘯 𝘕𝘦𝘵𝘸𝘰𝘳𝘬𝘴. 🔹 𝗕𝗿𝗼𝗮𝗱𝗲𝗿 𝗜𝗺𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻𝘀 𝗳𝗼𝗿 𝗘𝗺𝗯𝗼𝗱𝗶𝗲𝗱 𝗔𝗜 𝗮𝗻𝗱 𝗥𝗼𝗯𝗼𝘁𝗶𝗰𝘀: Cross-domain transfer methods for robotics and drones, to Embodied AI, and more. 🔹 𝗖𝗿𝗶𝘁𝗶𝗰𝗮𝗹 𝗖𝗵𝗮𝗹𝗹𝗲𝗻𝗴𝗲𝘀 𝗮𝗻𝗱 𝗢𝗽𝗽𝗼𝗿𝘁𝘂𝗻𝗶𝘁𝗶𝗲𝘀: generalization, safety, ethics, environment, and regulation. Strategic research paths in theoretical assurances, trust metrics, V2X cooperation, and urban integration. 🔹 𝗙𝘂𝘁𝘂𝗿𝗲 𝗢𝘂𝘁𝗹𝗼𝗼𝗸 𝗮𝗻𝗱 𝗩𝗶𝘀𝗶𝗼𝗻: dataset diversity, model robustness, theoretical foundations, computational scalability, human-AI collaborative frameworks. 📖 𝗥𝗲𝗮𝗱 𝘁𝗵𝗲 𝘀𝘂𝗿𝘃𝗲𝘆: arxiv.org/abs/2505.08854 🔗 E𝘅𝗽𝗹𝗼𝗿𝗲 𝗰𝗼𝗱𝗲: lnkd.in/gvcZh3Ka 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗔𝗜 𝗶𝘀 𝗳𝘂𝗻𝗱𝗮𝗺𝗲𝗻𝘁𝗮𝗹𝗹𝘆 𝗿𝗲𝗱𝗲𝗳𝗶𝗻𝗶𝗻𝗴 𝗳𝘂𝘁𝘂𝗿𝗲 𝘁𝗿𝗮𝗻𝘀𝗽𝗼𝗿𝘁𝗮𝘁𝗶𝗼𝗻. Our detailed survey equips 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵𝗲𝗿𝘀, 𝗲𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝘀, 𝗶𝗻𝗱𝘂𝘀𝘁𝗿𝘆 𝗹𝗲𝗮𝗱𝗲𝗿𝘀, 𝗮𝗻𝗱 𝗽𝗼𝗹𝗶𝗰𝘆𝗺𝗮𝗸𝗲𝗿𝘀 with the tools to navigate this transformative intersection. 🌟 The future of mobility is 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲. Let’s build it together! 🌟 #GenAI #AutonomusDriving #Autonomy #Drone #USDOT #USA #Diffusion #LLM #VLM #Collaboration #Research #Innovation #AI #CVPR #Publication #Overview