现已上线！2025 年 Telegram 研究 — 年度关键洞察 
AI with Papers - Artificial Intelligence & Deep Learning
前往频道在 Telegram
All the AI with papers. Every day fresh updates about #DeepLearning #MachineLearning #LLM & #ComputerVision Curated by Alessandro Ferrari | https://www.linkedin.com/in/visionarynet/ #AI #chatGPT
显示更多📈 Telegram 频道 AI with Papers - Artificial Intelligence & Deep Learning 的分析概览
频道 AI with Papers - Artificial Intelligence & Deep Learning (@ai_deeplearning) 英语 语言赛道中的 是活跃参与者。目前社区聚集了 17 166 名订阅者，在 技术与应用 类别中位列第 7 718，并在 马来西亚 地区排名第 2 234 位。
📊 受众指标与增长动态
自 невідомо 创建以来，项目保持高速增长，吸引了 17 166 名订阅者。
根据 20 六月, 2026 的最新数据，频道保持稳定运转。过去 30 天订阅人数变化为 -169，过去 24 小时变化为 0，整体触达仍然可观。
- 认证状态： 未认证
- 互动率 (ER)： 平均受众互动率为 22.86%。内容发布后 24 小时内通常能获得 N/A% 的反应，占订阅者总量。
- 帖子覆盖： 每篇帖子平均可获得 3 926 次浏览，首日通常累积 0 次浏览。
- 互动与反馈： 受众积极参与，单帖平均反应数为 26。
- 主题关注点： 内容集中在 framework, object, dataset, tba, depth 等核心主题上。
📝 描述与内容策略
作者将该频道定位为表达主观观点的平台：
“All the AI with papers. Every day fresh updates about #DeepLearning #MachineLearning #LLM & #ComputerVision
Curated by Alessandro Ferrari | https://www.linkedin.com/in/visionarynet/
#AI #chatGPT”
凭借高频更新（最新数据采集于 21 六月, 2026），频道始终保持新鲜度与高覆盖。分析显示受众积极互动，使其成为 技术与应用 类别中的关键影响点。
17 166
订阅者
无数据24 小时
-357 天
-16930 天
帖子存档
🫠 X-Portrait 2: SOTA(?) Portrait Animation 🫠
👉ByteDance unveils a preview of X-Portrait2, the new SOTA expression encoder model that implicitly encodes every minuscule expressions from the input by training it on large-scale datasets. Impressive results but no paper & code announced.
👉Review https://t.ly/8Owh9 [UPDATE]
👉Paper ?
👉Project byteaigc.github.io/X-Portrait2/
👉Repo ?
🧠 Single Neuron Reconstruction 🧠
👉SIAT unveils NeuroFly, a framework for large-scale single neuron reconstruction. Formulating neuron reconstruction task as a 3-stage streamlined workflow: automatic segmentation - connection - manual proofreading. Bridging computer vision and neuroscience 💙
👉Review https://t.ly/Y5Xu0
👉Paper https://arxiv.org/pdf/2411.04715
👉Repo github.com/beanli161514/neurofly
💪 Muscles in Time Dataset 💪
👉Muscles in Time (MinT) is a large-scale synthetic muscle activation dataset. MinT contains 9+ hours of simulation data covering 227 subjects and 402 simulated muscle strands. Code & Dataset available soon 💙
👉Review https://t.ly/108g6
👉Paper arxiv.org/pdf/2411.00128
👉Project davidschneider.ai/mint
👉Code github.com/simplexsigil/MusclesInTime
🏣 CityGaussianV2: Large-Scale City 🏣
👉A novel approach for large-scale scene reconstruction that addresses critical challenges related to geometric accuracy and efficiency: 10x compression, 25% faster & -50% memory! Source code released💙
👉Review https://t.ly/Xgn59
👉Paper arxiv.org/pdf/2411.00771
👉Project dekuliutesla.github.io/CityGaussianV2/
👉Code github.com/DekuLiuTesla/CityGaussian
☀️ Universal Relightable Avatars ☀️
👉#Meta unveils URAvatar, photorealistic & relightable avatars from phone scan with unknown illumination. Stunning results!
👉Review https://t.ly/U-ESX
👉Paper arxiv.org/pdf/2410.24223
👉Project junxuan-li.github.io/urgca-website
☀️ Universal Relightable Avatars ☀️
👉#Meta unveils URAvatar, photorealistic & relightable avatars from phone scan with unknown illumination. Stunning results!
👉Review https://t.ly/U-ESX
👉Paper arxiv.org/pdf/2410.24223
👉Project junxuan-li.github.io/urgca-website
🍜 REM: Segment What You Describe 🍜
👉REM is a framework for segmenting concepts in video that can be described via LLM. Suitable for rare & non-object dynamic concepts, such as waves, smoke, etc. Code & Data announced 💙
👉Review https://t.ly/OyVtV
👉Paper arxiv.org/pdf/2410.23287
👉Project https://miccooper9.github.io/projects/ReferEverything/
🔥🔥 The code is out 🔥🔥
👉Code https://github.com/HaixinShi/fmov_pose
🔥 D-FINE: new SOTA Detector 🔥
👉D-FINE, a powerful real-time object detector that achieves outstanding localization precision by redefining the bounding box regression task in DETR model. New SOTA on MS COCO with additional data. Code & models available 💙
👉Review https://t.ly/aw9fN
👉Paper https://arxiv.org/pdf/2410.13842
👉Code https://github.com/Peterande/D-FINE
🫐 Blendify: #Python + Blender 🫐
👉Lightweight Python framework that provides a high-level API for creating & rendering scenes with #Blender. It simplifies data augmentation & synthesis. Source Code released💙
👉Review https://t.ly/l0crA
👉Paper https://arxiv.org/pdf/2410.17858
👉Code https://virtualhumans.mpi-inf.mpg.de/blendify/
⛈️ SMITE: SEGMENT IN TIME ⛈️
👉SFU unveil SMITE: a novel AI that -with only one or few segmentation references with fine granularity- is able to segment different unseen videos respecting the segmentation references. Dataset & Code (under Apache 2.0) announced 💙
👉Review https://t.ly/w6aWJ
👉Paper arxiv.org/pdf/2410.18538
👉Project segment-me-in-time.github.io/
👉Code github.com/alimohammadiamirhossein/smite/
🌻 Plant Camouflage Detection🌻
👉PlantCamo Dataset is the first dataset for plant camouflage detection: 1,250 images with camouflage characteristics. Source Code released 💙
👉Review https://t.ly/pYFX4
👉Paper arxiv.org/pdf/2410.17598
👉Code github.com/yjybuaa/PlantCamo
🪁 PL2Map: efficient neural 2D-3D 🪁
👉PL2Map is a novel neural network tailored for efficient representation of complex point & line maps. A natural representation of 2D-3D correspondences
👉Review https://t.ly/D-bVD
👉Paper arxiv.org/pdf/2402.18011
👉Project https://thpjp.github.io/pl2map
👉Code https://github.com/ais-lab/pl2map
🧿 Look Ma, no markers 🧿
👉#Microsoft unveils the first technique for marker-free, HQ reconstruction of COMPLETE human body, including eyes and tongue, without requiring any calibration, manual intervention or custom hardware. Impressive results! Repo for training & Dataset released💙
👉Review https://t.ly/5fN0g
👉Paper arxiv.org/pdf/2410.11520
👉Project microsoft.github.io/SynthMoCap/
👉Repo github.com/microsoft/SynthMoCap
🔥BitNet: code of 1-bit LLM is out 🔥
👉BitNet by #Microsoft, announced in late 2023, is a 1-bit Transformer architecture designed for LLMs. BitLinear as a drop-in replacement of the nn.Linear layer in order to train 1-bit weights from scratch. Source Code just released a few hours ago 💙
👉Review https://t.ly/3G2LA
👉Paper arxiv.org/pdf/2310.11453
👉Code https://lnkd.in/duPADJVb
☀️ GS + Depth = SOTA ☀️
👉ETH unveils DepthSplat, the new SOTA in depth estimation and novel view synthesis tasks. The key feature is the cross-task interactions between Gaussian Splatting & depth estimation. Source Code to be released in a few days💙
👉Review https://t.ly/87HuH
👉Paper arxiv.org/abs/2410.13862
👉Project haofeixu.github.io/depthsplat/
👉Code github.com/cvg/depthsplat
🦠 Neural Metamorphosis 🦠
👉NU Singapore unveils NeuMeta to transform neural nets by allowing a single model to adapt on the fly to different sizes, generating the right weights when needed.
👉Review https://t.ly/DJab3
👉Paper arxiv.org/pdf/2410.11878
👉Project adamdad.github.io/neumeta
👉Code github.com/Adamdad/neumeta
🔥 CoTracker3 by #META is out! 🔥
👉#Meta (+VGG Oxford) unveils CoTracker3, a new tracker that outperforms the previous SoTA by a large margin using only the 0.1% of the training data 🤯🤯🤯
👉Review https://t.ly/TcRIv
👉Paper arxiv.org/pdf/2410.11831
👉Project cotracker3.github.io/
👉Code github.com/facebookresearch/co-tracker
🪞Robo-Emulation via Video Imitation🪞
👉OKAMI (UT & #Nvidia) is a novel foundation method that generates a manipulation plan from a single RGB-D video and derives a policy for execution.
👉Review https://t.ly/_N29-
👉Paper arxiv.org/pdf/2410.11792
👉Project https://lnkd.in/d6bHF_-s
🔥 DEPTH ANY VIDEO is out! 🔥
👉DAV is a novel foundation model for image/video depth estimation.The new SOTA for accuracy & consistency, up to 150 FPS!
👉Review https://t.ly/CjSz2
👉Paper arxiv.org/pdf/2410.10815
👉Project depthanyvideo.github.io/
👉Code github.com/Nightmare-n/DepthAnyVideo
