[DL輪読会]A Probabilistic U-Net for Segmentation of Ambiguous Images

Oct 26, 20185 likes1,370 views

This paper proposes a probabilistic U-Net model for image segmentation that models segmentation uncertainty. The model consists of a prior network that predicts an initial segmentation map and a posterior network that refines the prior map based on the input image. The loss function maximizes the posterior probability of the ground truth segmentation while minimizing the divergence between the prior and posterior predictions, allowing the model to capture ambiguous regions. Experiments on medical images show the probabilistic U-Net produces better segmentations and uncertainty estimates than deterministic segmentation models.

DEEP LEARNING JP
[DL Papers]
A Probabilistic U-Net for Segmentation of Ambiguous Images
Tomoki Tanimura, Keio University
http://deeplearning.jp/
1

RNS
§ .0
§ P 0 8: : ,
§ 81 1 2 1 2
§ , I MCDAG MC K

N , ,- ,
§
§ C : F
§ , IS R , R
§ M ,- , ,

C
§ + ( )
)
§ E 3
§ 1 -
§ Segmentation Variant6
§ U-Net z6 6 A

[DL輪読会]A Probabilistic U-Net for Segmentation of Ambiguous Images

- :
LossLoss
§ Posterior (Prior )
ν
ω
X
θ
ψ
z
S Y

§
§ C [ C R B
§
§ ( - .
§
§ ) ) )
§ D 4 E G I
§
§ LD T

] 2 2
§ 2 . 2 . epS v
§ s V tI E A
§ E xn gl
§ d h fa [ .2 I o ci
P Irm

U
§ (2 2 21 2 = : 2 21
§ Y a T
§ : (, : 21 21 dP (,
§ 1 = ) = E GI
§ S d S d cD

T
§ (2 2 21 2 = : 2 21
§ c Y
§ : (, a : 21 a 21 P (,
§ 1 = ) = E S G
§ Ud Ud eD
GT Pred
S
2I Pred
S
2I GT
S

GT2Pred 2 Pred 2 GT
§ 2 ) G
§ G
§ 2 ) ()
G 1
§

§
§ d 1/: 2/ /
§ e 1/: af G/:: / 7 : 4
§ /7:7:5 ./972/ 7 :
§ -
§ b
§
§ 3 2
§ V V c gGT ,
§ 72 /98C 72 /98 C 7 6 / / 797 4
§ :C : C 7 6 / / 797 4
§ 1/ C 1/ C 7 6
§ 5 / 7 :C 5 / 7 : C 7 6
§ /2C /2 C 7 6 / 797

g
§
§
§ n D
§
§ d 4 D T
§ m 3 e D i
§ S D 3
§ a 3 2 3
§ o 3 G

a
§ , ,6 14 8 - bS
§ PD N a U
§
§ e ) - (
§ N 4 , 1 4 d

e a -
§ ) cbI P :I
§ N ) T
§ ( - T
§ ( 1 U ) I ( C
§ )C( T d ( 1
U G

N S o
iT r
Prior m P n fix
V Posterior GT
Prior U-Net
§ iT r Gt
§ - e n V
§ -
Ug
§ -
a

§ N Gi - - + S P pm
§ s n E N g N - C
- U r
§ t + n - - + U S TV + -
o e ec
§ r ya e TV AC

The document discusses Deep Learning Japan (DL Papers), a website that aggregates and shares Japanese-language papers on deep learning. It provides an overview of the website's features and content, including sections on recent papers, tutorials, tools and frameworks. In summary: - DL Papers collects and shares Japanese papers on deep learning techniques to help disseminate research. - The site organizes papers into categories like recent publications, tutorials, tools and frameworks. - It aims to help more researchers engage with deep learning and accelerate progress in the field through open sharing of ideas.

異常検知とGAN: AnoGanKoichiro tamura

DNNコンパイラの歩みと最近の動向〜TVMを中心に〜Takeo Imai

2019.2.2 fpgax #11 ＋ TFUG ハード部：DNN専用ハードについて語る会発表資料 AIチップの登場やエッジデバイスでのAI/DL推論へのニーズが高まるのに伴い、学習済みDNNを多様なハードウェアに合わせて最適化し実行コードに変換するDNNコンパイラが多数登場しました。本講演ではDNNコンパイラの成り立ちから基本的な構成、そして現在の技術開発動向について紹介します。

「解説資料」Toward Fast and Stabilized GAN Training for High-fidelity Few-shot Imag...Takumi Ohkuma

GPGPU Seminar (GPU Accelerated Libraries, 1 of 3, cuBLAS) 智啓出川

GPU最適化ライブラリの利用（1/3，cuBLAS）長岡技術科学大学2015年度GPGPU講習会（2015年10月14日実施） CUDA付属のライブラリを利用して連立一次方程式を解く内容を3回に分けて実施． 1回目　cuBLASを利用したGPU移植 2回目　cuSPARSEを利用した省メモリ化 http://www.slideshare.net/ssuserf87701/gpgpu-seminar-gpu-accelerated-libraries-2-of-3-cusparse 3回目　Thrustを利用した実装効率化 http://www.slideshare.net/ssuserf87701/gpgpu-seminar-gpu-accelerated-libraries-3-of-3-thrust 講義には長岡技術科学大学のGPGPUシステム（GROUSE）を利用しています。開発環境 CPU　Intel Xeon X5670 × 32 GPU　NVIDIA Tesla M2050（Fermi世代） × 64 CUDA 4.0（諸般の事情によりバージョンアップされていません。） PGI Fortran 11.3 (with CUDA 3.2) GPGPU講習会・PyCUDA http://www.slideshare.net/ssuserf87701/gpgpu-seminar-pycuda ・CUDA Fortranによる格子ボルツマン法の高速化 http://www.slideshare.net/ssuserf87701/gpgpu-seminar-accelerataion-of-lattice-boltzmann-method-using-cuda-fortran ・補足資料　GPGPUとCUDA Fortran http://www.slideshare.net/ssuserf87701/gpgpu-seminar-gpgpu-and-cuda-fortran ・GPU最適化ライブラリの利用（その１，cuBLAS） http://www.slideshare.net/ssuserf87701/gpgpu-seminar-gpu-accelerated-libraries-1-of-3-cublas ・GPU最適化ライブラリの利用（その２，cuSPARSE） http://www.slideshare.net/ssuserf87701/gpgpu-seminar-gpu-accelerated-libraries-2-of-3-cusparse ・GPU最適化ライブラリの利用（その３，Thrust） http://www.slideshare.net/ssuserf87701/gpgpu-seminar-gpu-accelerated-libraries-3-of-3-thrust 2015年度GPGPU実践基礎工学・第1回　学際的分野における先端シミュレーション技術の歴史 http://www.slideshare.net/ssuserf87701/2015gpgpu1 2015年度GPGPU実践プログラミング・第1回　GPGPUの歴史と応用例 http://www.slideshare.net/ssuserf87701/2015gpgpu1-59179080 2015年度先端GPGPUシミュレーション工学特論・第1回　先端シミュレーションおよび産業界におけるGPUの役割 http://www.slideshare.net/ssuserf87701/2015gpgpu1-59180313

[DL輪読会]Weight Agnostic Neural NetworksDeep Learning JP

【DL輪読会】Unbiased Gradient Estimation for Marginal Log-likelihoodDeep Learning JP

1. The document proposes methods for estimating the marginal log-likelihood of latent variable models in an unbiased manner. 2. It discusses using Monte Carlo methods like MCMC and importance sampling to estimate the intractable integral in the marginal log-likelihood. Multilevel Monte Carlo can provide an unbiased estimate with fewer samples than standard Monte Carlo. 3. Stochastically Unbiased Marginalization Objective (SUMO) is introduced to provide an unbiased estimate of the marginal log-likelihood using a single sample. This involves weighting the importance weighted bound with a geometric distribution.

Improving Language Understanding by Generative Pre-TrainingMasao Taketani

【DL輪読会】論文解説：Offline Reinforcement Learning as One Big Sequence Modeling ProblemDeep Learning JP

[DL Hacks]Simple Online Realtime Tracking with a Deep Association MetricDeep Learning JP

物体検出の歴史（R-CNNからSSD・YOLOまで）HironoriKanazawa

Outrageously Large Neural Networks:The Sparsely-Gated Mixture-of-Experts Laye...KCS Keio Computer Society

SSII2019TS: 実践カメラキャリブレーション～カメラを用いた実世界計測の基礎と応用～SSII

SSII2019 チュートリアル講演 TS3 6月14日(金) 9:00～10:20 (メインホール) カメラの幾何学的キャリブレーションはカメラを使った３次元計測・認識に欠かすことができない基礎技術です。近年では LIDAR や ToF カメラのように深度情報を直接出力するカメラや、全周囲カメラのように透視投影ではないカメラをセンサとして使用することも一般的になりました。本講演ではこれらの幾何学的キャリブレーションについて、実例を交えながら基礎からご紹介します。さらに講演内容に対応した初学者向けのサンプルコードについても紹介します。

[DL輪読会]GQNと関連研究，世界モデルとの関係についてDeep Learning JP

[DLHacks 実装] DeepPose: Human Pose Estimation via Deep Neural NetworksDeep Learning JP

【DL輪読会】Diffusion Policy: Visuomotor Policy Learning via Action DiffusionDeep Learning JP

This document discusses a paper on visuomotor policy learning via action diffusion. The paper presents a method for training policies that map camera images directly to actions by incorporating action diffusion, which adds noise to actions during training. This helps explore the action space and avoid getting stuck in local optima during policy learning. The method can learn policies for complex manipulation tasks entirely from pixels using self-supervised reinforcement learning with image rewards.

【DL輪読会】GET3D: A Generative Model of High Quality 3D Textured Shapes Learned f...Deep Learning JP

R Markdownによるドキュメント生成とバージョン管理入門nocchi_airport

SSII2022 [SS2] 少ないデータやラベルを効率的に活用する機械学習技術〜足りない情報をどのように補うか？〜SSII

6/9 (木) 14:45～15:15 メイン会場講師：石井雅人氏（ソニーグループ株式会社）概要：機械学習技術の急速な発達により、コンピュータによる知的処理は様々なタスクで人間に匹敵あるいは凌駕する性能を達成してきた。一方、このような高い性能は大量かつ高品質な学習データによって支えられており、多様化する機械学習応用においてデータの収集コストが大きな導入障壁の１つとなっている。本講演では、少ないデータやラベルから効率的に学習するための様々な技術について、「足りない情報をどのように補うか？」という観点から概観するとともに、特に画像認識分野における最新の研究動向についても紹介する。

Action Recognitionの歴史と最新動向Ohnishi Katsunori

This document summarizes recent developments in action recognition using deep learning techniques. It discusses early approaches using improved dense trajectories and two-stream convolutional neural networks. It then focuses on advances using 3D convolutional networks, enabled by large video datasets like Kinetics. State-of-the-art results are achieved using inflated 3D convolutional networks and temporal aggregation methods like temporal linear encoding. The document provides an overview of popular datasets and challenges and concludes with tips on training models at scale.

StyleGAN解説 CVPR2019読み会@DeNAKento Doi

Swin Transformer (ICCV'21 Best Paper) を完璧に理解する資料Yusuke Uchida

分散深層学習 @ NIPS'17Takuya Akiba

BERT入門Ken'ichi Matsui

【DL輪読会】Hierarchical Text-Conditional Image Generation with CLIP LatentsDeep Learning JP

生成モデルの Deep LearningSeiya Tokui

第 1 回 Jetson ユーザー勉強会NVIDIA Japan

この資料は 5 月 24 日 (金) にソフトバンク新館汐留キャンパスにて開催の「第 1 回 Jetson ユーザー勉強会」https://jsjug.connpass.com/event/126529/ における、NVIDIA の資料です。＜アジェンダ＞ ➢ Jetson Platform のご紹介 • Jetson 採用事例 • Jetson ファミリラインアップ • Jetson Nano の概要 ➢ Jetson Nano Getting Started • 各種のドキュメント、リソース • システムセットアップ • 電力効率 & パフォーマンスモニター • GPIO & アクセサリ ➢ DeepStream と TensorRT の概要 ➢ Jetson Nano による DeepStream デモンストレーション

【メタサーベイ】Vision and Language のトップ研究室/研究者cvpaper. challenge

Kaggle Google Quest Q&A Labeling 反省会 LT資料 47th place solutionKen'ichi Matsui

The document discusses different approaches that were tried for improving the performance of a model for a question answering competition, including pre-training on additional data, modifying the model architecture by changing layers or heads, and using different loss functions or features. Various models were experimented with, such as BERT, RoBERTa, ALBERT, and XLNet. However, concatenating the question and answer encodings did not work as expected.

DevOps導入支援サービスArata Fujimura

More Related Content

What's hot (20)

【DL輪読会】論文解説：Offline Reinforcement Learning as One Big Sequence Modeling ProblemDeep Learning JP

[DL Hacks]Simple Online Realtime Tracking with a Deep Association MetricDeep Learning JP

物体検出の歴史（R-CNNからSSD・YOLOまで）HironoriKanazawa

Outrageously Large Neural Networks:The Sparsely-Gated Mixture-of-Experts Laye...KCS Keio Computer Society

SSII2019TS: 実践カメラキャリブレーション～カメラを用いた実世界計測の基礎と応用～SSII

[DL輪読会]GQNと関連研究，世界モデルとの関係についてDeep Learning JP

[DLHacks 実装] DeepPose: Human Pose Estimation via Deep Neural NetworksDeep Learning JP

【DL輪読会】Diffusion Policy: Visuomotor Policy Learning via Action DiffusionDeep Learning JP

【DL輪読会】GET3D: A Generative Model of High Quality 3D Textured Shapes Learned f...Deep Learning JP

R Markdownによるドキュメント生成とバージョン管理入門nocchi_airport

SSII2022 [SS2] 少ないデータやラベルを効率的に活用する機械学習技術〜足りない情報をどのように補うか？〜SSII

Action Recognitionの歴史と最新動向Ohnishi Katsunori

StyleGAN解説 CVPR2019読み会@DeNAKento Doi

Swin Transformer (ICCV'21 Best Paper) を完璧に理解する資料Yusuke Uchida

分散深層学習 @ NIPS'17Takuya Akiba

BERT入門Ken'ichi Matsui

【DL輪読会】Hierarchical Text-Conditional Image Generation with CLIP LatentsDeep Learning JP

生成モデルの Deep LearningSeiya Tokui

第 1 回 Jetson ユーザー勉強会NVIDIA Japan

【メタサーベイ】Vision and Language のトップ研究室/研究者cvpaper. challenge

【DL輪読会】論文解説：Offline Reinforcement Learning as One Big Sequence Modeling ProblemDeep Learning JP

[DL Hacks]Simple Online Realtime Tracking with a Deep Association MetricDeep Learning JP

物体検出の歴史（R-CNNからSSD・YOLOまで）HironoriKanazawa

Outrageously Large Neural Networks:The Sparsely-Gated Mixture-of-Experts Laye...KCS Keio Computer Society

SSII2019TS: 実践カメラキャリブレーション～カメラを用いた実世界計測の基礎と応用～SSII

[DL輪読会]GQNと関連研究，世界モデルとの関係についてDeep Learning JP

[DLHacks 実装] DeepPose: Human Pose Estimation via Deep Neural NetworksDeep Learning JP

【DL輪読会】Diffusion Policy: Visuomotor Policy Learning via Action DiffusionDeep Learning JP

【DL輪読会】GET3D: A Generative Model of High Quality 3D Textured Shapes Learned f...Deep Learning JP

R Markdownによるドキュメント生成とバージョン管理入門nocchi_airport

SSII2022 [SS2] 少ないデータやラベルを効率的に活用する機械学習技術〜足りない情報をどのように補うか？〜SSII

Action Recognitionの歴史と最新動向Ohnishi Katsunori

StyleGAN解説 CVPR2019読み会@DeNAKento Doi

Swin Transformer (ICCV'21 Best Paper) を完璧に理解する資料Yusuke Uchida

分散深層学習 @ NIPS'17Takuya Akiba

BERT入門Ken'ichi Matsui

【DL輪読会】Hierarchical Text-Conditional Image Generation with CLIP LatentsDeep Learning JP

生成モデルの Deep LearningSeiya Tokui

第 1 回 Jetson ユーザー勉強会NVIDIA Japan

【メタサーベイ】Vision and Language のトップ研究室/研究者cvpaper. challenge

Similar to [DL輪読会]A Probabilistic U-Net for Segmentation of Ambiguous Images (20)

Kaggle Google Quest Q&A Labeling 反省会 LT資料 47th place solutionKen'ichi Matsui

DevOps導入支援サービスArata Fujimura

Attention-Based Adaptive Selection of Operations for Image Restoration in the...MasanoriSuganuma

The document appears to be a technical paper describing a machine learning model for image restoration. It presents results of the proposed model on several benchmark datasets compared to other methods. The model uses an operation-wise attention mechanism to identify and restore corrupted image regions. Results show the method works well for various types of image distortions and outperforms previous single image restoration approaches.

DevOps導入支援サービス(Ver.2)Arata Fujimura

Google Cloud Dataflowによるデータ変換処理入門Takeshi Mikami

Continuous Delivery with Spinnaker on K8s(kubernetes) Cluster Jinwoong Kim

This document discusses access control and authentication for a system. It covers setting up user accounts with different permissions, including setting permissions for specific directories and files. Authentication is discussed, including using passwords and ensuring they meet strength requirements. Auditing is also mentioned to log access and changes to files and directories. The document provides guidance on setting up and configuring this access control and authentication system.

MAGO3Dによる３次元マルチスケールデータの可視化システムの開発(FOSS4G_TOKAI_2018)Hirofumi Hayashi

Distributed Denial of Service Attack Prevention at Source MachinesShinagawa Laboratory, The University of Tokyo

[DL輪読会]深層強化学習はなぜ難しいのか？Why Deep RL fails? A brief survey of recent works.Deep Learning JP

Deep reinforcement learning algorithms often fail to learn complex tasks. Recent works have identified three issues that form a "deadly triad" contributing to this problem: non-stationary targets, high variance, and positive correlation. New algorithms aim to address these issues by improving exploration, stabilizing learning, and decorrelating updates. Overall, deep reinforcement learning remains a challenging area with opportunities to develop more data-efficient and generally applicable algorithms.

リアルタイム議事録＆翻訳付きのビデオ会議を作ろう ~WebRTCの最新動向~ SkyWay Media Pipeline FactoryRyosuke Otsuya

APPS JAPAN 2019 専門セミナー 2019/06/12 リアルタイム議事録&翻訳付きのビデオ会議を作ろう～WebRTCの最新動向～ NTTコミュニケーションズ技術開発部担当課長大津谷亮祐 WebRTCを使えば、WebサイトやiOS・Androidアプリに、手軽にビデオ通話を組み込むことができます。最近では、WebRTCと音声認識などのクラウドAIを組み合わせる事例も出てきました。このようなWebRTCの最新動向をご紹介し、一例として、SkyWay Media Pipeline Factoryという無料サービスを使って、リアルタイム議事録&翻訳付きのビデオ会議を作る方法を解説します。

第二次政策提言抜粋版 2019.3.22立憲パートナーズ社会構想研究会（非公式）

This document contains a series of codes and symbols that appear to be encrypting or obfuscating text-based messages. Some of the codes include numbers, punctuation marks, and repeating patterns that do not form clear language on their own. The length and format of the document suggests it is conveying multiple short encrypted messages but the exact meaning is not discernible without the decryption key.

Kubernetes ネットワーキングのすべてLINE Corporation

Db2 Warehouse v3.0 SMP 導入ガイド 20190104 Db2 Warehouse SMP v3.0 configration Ins...IBM Analytics Japan

Db2 Warehouseは分析用途に最適化されたDockerコンテナで提供されるDWHソフトウェアアプライアンスです。高速性、簡易性、柔軟性に優れた特長を持つプライベートクラウド向けデータ分析基盤です。本ガイドはDb2 Warehouse SMP(シングルノード構成)の導入手順書です。 2 Warehouse is a DWH software appliance offered in a Docker container optimized for analysis purposes. It is a data analysis infrastructure for private clouds with superior features of high speed, simplicity and flexibility. This guide is the installation procedure of Db 2 Warehouse SMP (single node configuration).

[DL Hacks 実装]Attention is All You NeedDeep Learning JP

1. The document discusses artificial intelligence and machine learning techniques including neural networks, self-attention mechanisms, and recurrent neural network grammars. 2. Key applications mentioned include natural language processing tasks like machine translation, summarization, and question answering as well as generating realistic fake text. 3. Challenges and open problems are also outlined, such as how to ensure AI systems are robust, safe, and aligned with human values and priorities.

Spark MLlib ML Pipelines の概要及びpysparkからの扱い方Takeshi Mikami

OSC 2018 Nagoya rsyncやシェルでバックアップするよりも簡単にOSSのBaculaでバックアップしてみようKen Sawada

Towards formal verification of neural networksMasahiro Sakai

Edge trends mizunoshintaro mizuno

- Edge computing enables applications and services to be located closer to users and devices, reducing latency and network congestion. This provides for new use cases and improved user experiences. - The document discusses edge reference architectures including small, medium, and large edge deployments. It also covers considerations for connectivity, compute, storage and management at the edge. - Openstack is working to develop standards and support edge computing through projects like the Edge Computing Group to help drive open edge infrastructure.

вестник южно уральского-государственного_университета._серия_математика._меха...Иван Иванов

- The article deals with surfaces of negative Gaussian curvature that can be bijectively projected onto a circle. - The author provides sufficient conditions for the existence of an estimate of the circle radius onto which the surface can be projected. - Specifically, if the Gaussian curvature is bounded above by a negative constant, an estimate of the minimum possible radius of the projecting circle can be determined.

【CVPR 2019】Learning Cross Modal Embeddings with Adversarial Networks for Cook...cvpaper. challenge

cvpaper.challenge はコンピュータビジョン分野の今を映し、トレンドを創り出す挑戦です。論文読破・まとめ・アイディア考案・議論・実装・論文投稿に取り組み、あらゆる知識を共有しています。 http://xpaperchallenge.org/cv/ 本資料は、CVPR 2019 網羅的サーベイの成果の一部で、1論文を精読してプレゼンテーション形式でまとめております。論文サマリは下記からご確認頂けます。 http://xpaperchallenge.org/cv/survey/cvpr2019_summaries/listall/

Kaggle Google Quest Q&A Labeling 反省会 LT資料 47th place solutionKen'ichi Matsui

DevOps導入支援サービスArata Fujimura

Attention-Based Adaptive Selection of Operations for Image Restoration in the...MasanoriSuganuma

DevOps導入支援サービス(Ver.2)Arata Fujimura

Google Cloud Dataflowによるデータ変換処理入門Takeshi Mikami

Continuous Delivery with Spinnaker on K8s(kubernetes) Cluster Jinwoong Kim

MAGO3Dによる３次元マルチスケールデータの可視化システムの開発(FOSS4G_TOKAI_2018)Hirofumi Hayashi

Distributed Denial of Service Attack Prevention at Source MachinesShinagawa Laboratory, The University of Tokyo

[DL輪読会]深層強化学習はなぜ難しいのか？Why Deep RL fails? A brief survey of recent works.Deep Learning JP

リアルタイム議事録＆翻訳付きのビデオ会議を作ろう ~WebRTCの最新動向~ SkyWay Media Pipeline FactoryRyosuke Otsuya

第二次政策提言抜粋版 2019.3.22立憲パートナーズ社会構想研究会（非公式）

Kubernetes ネットワーキングのすべてLINE Corporation

Db2 Warehouse v3.0 SMP 導入ガイド 20190104 Db2 Warehouse SMP v3.0 configration Ins...IBM Analytics Japan

[DL Hacks 実装]Attention is All You NeedDeep Learning JP

Spark MLlib ML Pipelines の概要及びpysparkからの扱い方Takeshi Mikami

OSC 2018 Nagoya rsyncやシェルでバックアップするよりも簡単にOSSのBaculaでバックアップしてみようKen Sawada

Towards formal verification of neural networksMasahiro Sakai

Edge trends mizunoshintaro mizuno

вестник южно уральского-государственного_университета._серия_математика._меха...Иван Иванов

【CVPR 2019】Learning Cross Modal Embeddings with Adversarial Networks for Cook...cvpaper. challenge

More from Deep Learning JP (20)

【DL輪読会】AdaptDiffuser: Diffusion Models as Adaptive Self-evolving PlannersDeep Learning JP

【DL輪読会】事前学習用データセットについてDeep Learning JP

【DL輪読会】 "Learning to render novel views from wide-baseline stereo pairs." CVP...Deep Learning JP

【DL輪読会】Zero-Shot Dual-Lens Super-ResolutionDeep Learning JP

【DL輪読会】BloombergGPT: A Large Language Model for Finance arxivDeep Learning JP

【DL輪読会】マルチモーダル LLMDeep Learning JP

【 DL輪読会】ToolLLM: Facilitating Large Language Models to Master 16000+ Real-wo...Deep Learning JP

【DL輪読会】AnyLoc: Towards Universal Visual Place RecognitionDeep Learning JP

【DL輪読会】Can Neural Network Memorization Be Localized?Deep Learning JP

【DL輪読会】Hopfield network　関連研究についてDeep Learning JP

【DL輪読会】SimPer: Simple self-supervised learning of periodic targets( ICLR 2023 )Deep Learning JP

【DL輪読会】RLCD: Reinforcement Learning from Contrast Distillation for Language M...Deep Learning JP

【DL輪読会】"Secrets of RLHF in Large Language Models Part I: PPO"Deep Learning JP

【DL輪読会】"Language Instructed Reinforcement Learning for Human-AI Coordination "Deep Learning JP

【DL輪読会】Llama 2: Open Foundation and Fine-Tuned Chat ModelsDeep Learning JP

【DL輪読会】"Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware"Deep Learning JP

【DL輪読会】Parameter is Not All You Need:Starting from Non-Parametric Networks fo...Deep Learning JP

【DL輪読会】Drag Your GAN: Interactive Point-based Manipulation on the Generative ...Deep Learning JP

【DL輪読会】Self-Supervised Learning from Images with a Joint-Embedding Predictive...Deep Learning JP

【DL輪読会】Towards Understanding Ensemble, Knowledge Distillation and Self-Distil...Deep Learning JP