2024.11.26 每日AI论文 | 3D材料生成自动化,零样本图像生成创新。
Listen now
Description
本期的 21 篇论文如下: [00:26] 🌐 Material Anything: Generating Materials for Any 3D Object via Diffusion(材料生成:通过扩散生成任意3D对象的材料) [01:05] 🎨 Large-Scale Text-to-Image Model with Inpainting is a Zero-Shot Subject-Driven Image Generator(基于修复的大规模文本到图像模型:零样本主题驱动图像生成器) [01:48] 🤖 From Generation to Judgment: Opportunities and Challenges of LLM-as-a-judge(从生成到判断:LLM作为评判者的机遇与挑战) [02:22] 🌐 Knowledge Transfer Across Modalities with Natural Language Supervision(基于自然语言监督的多模态知识迁移) [03:00] 🧠 MH-MoE:Multi-Head Mixture-of-Experts(多头混合专家模型) [03:34] 🎥 DreamRunner: Fine-Grained Storytelling Video Generation with Retrieval-Augmented Motion Adaptation(DreamRunner:基于检索增强的运动适应细粒度故事视频生成) [04:13] 🌐 One Diffusion to Generate Them All(一个扩散模型生成所有) [04:54] 👁 VisualLens: Personalization through Visual History(视觉透镜:通过视觉历史实现个性化) [05:34] 🔍 Factorized Visual Tokenization and Generation(因子分解视觉标记化与生成) [06:15] 🔍 O1 Replication Journey -- Part 2: Surpassing O1-preview through Simple Distillation, Big Progress or Bitter Lesson?(O1复制之旅 -- 第二部分:通过简单蒸馏超越O1预览版,巨大进步还是苦涩教训?) [07:00] 🩺 GMAI-VL & GMAI-VL-5.5M: A Large Vision-Language Model and A Comprehensive Multimodal Dataset Towards General Medical AI(通用医疗人工智能的大规模视觉语言模型与综合多模态数据集) [07:39] 🌐 SplatFlow: Multi-View Rectified Flow Model for 3D Gaussian Splatting Synthesis(SplatFlow:用于3D高斯喷射合成的多视图校正流模型) [08:25] 🔄 From CISC to RISC: language-model guided assembly transpilation(从CISC到RISC:语言模型引导的汇编转译) [09:03] ⚙ Cautious Optimizers: Improving Training with One Line of Code(谨慎优化器:用一行代码改进训练) [09:49] 🤖 The Impossible Test: A 2024 Unsolvable Dataset and A Chance for an AGI Quiz(不可能的测试:2024年不可解数据集与AGI测验的机会) [10:30] 🔮 Predicting Emergent Capabilities by Finetuning(通过微调预测涌现能力) [11:04] 📊 SegBook: A Simple Baseline and Cookbook for Volumetric Medical Image Segmentation(SegBook:体积医学图像分割的简单基线和操作手册) [11:48] 🩺 Interactive Medical Image Segmentation: A Benchmark Dataset and Baseline(交互式医学图像分割:基准数据集与基线) [12:25] 🤔 LLMs Do Not Think Step-by-step In Implicit Reasoning(大语言模型在隐式推理中不进行逐步思考) [13:00] 🌐 Best of Both Worlds: Advantages of Hybrid Graph Sequence Models(双剑合璧:混合图序列模型的优势) [13:34] 🔗 Edge Weight Prediction For Category-Agnostic Pose Estimation(类别无关姿态估计的边权重预测) 【关注我们】 您还可以在以下平台找到我们,获得播客内容以外更多信息 小红书: AI速递
More Episodes
本期的 18 篇论文如下: [00:28] 🖥 ShowUI: One Vision-Language-Action Model for GUI Visual Agent(ShowUI:一种用于GUI视觉代理的视觉-语言-动作模型) [01:08] 🎥 Pathways on the Image Manifold: Image Editing via Video Generation(图像流形上的路径:通过视频生成进行图像编辑) [01:45] ⭐ Star Attention: Efficient LLM Inference over Long...
Published 11/27/24
本期的 14 篇论文如下: [00:26] 🎨 Style-Friendly SNR Sampler for Style-Driven Generation(风格友好SNR采样器用于风格驱动生成) [01:08] 🚀 TÜLU 3: Pushing Frontiers in Open Language Model Post-Training(TÜLU 3:推动开放语言模型后训练的前沿) [01:53] 🌐 OminiControl: Minimal and Universal Control for Diffusion...
Published 11/25/24