PinnedMakoto TAKAMATSUinArtificial Intelligence in Plain EnglishImprove Segmentation Accuracy by Correcting for Gaps in PredictionsSilhouette mask segmentation mimics the clinician’s annotation approach5 min read·Dec 20, 2020----
PinnedMakoto TAKAMATSUinTowards AIApplication of Uncertainty Modeling to Face RecognitionAn facial recognition algorithm that effectively mitigates the negative impact of dirty samples during model training8 min read·Dec 3, 2020----
PinnedMakoto TAKAMATSU[Review] Poor Images Diagnosis [part1]Towards a reliable medical AI2 min read·Nov 19, 2020----
Makoto TAKAMATSUinAIメディカルサービスBottleneck TransformerSelf attentionをResNetに組み込んだTransformer7 min read·Jul 7, 2021----
Makoto TAKAMATSUinGeek CultureGastric Histopathology Image Classification by Transformer, GasHis-TransformerCNN and Transformer Fusion for Gastric Histopathology Image Classification with Few False Negatives and False Positive3 min read·Jun 21, 2021----
Makoto TAKAMATSUinAIメディカルサービスTransformerによる胃組織病理画像分類,GasHis-TransformerCNNとTransformerの融合で偽陰性の少ない胃組織病理学的画像分類を実現7 min read·Jun 20, 2021----
Makoto TAKAMATSUinAIメディカルサービス[Transformer &物体検知] You Only Look at One Sequence (YOLOS)画像識別タスクで事前学習したTransformerを物体検出タスクに転移学習ができる!?10 min read·Jun 9, 2021----
Makoto TAKAMATSUinAIメディカルサービス[論文紹介] GANを使用しない効果的なドメイン適応教師なしドメイン適応,ソース分布とターゲット分布間の不一致を低周波スペクトル交換で減少6 min read·May 26, 2021----
Makoto TAKAMATSUinTowards AIA Useful New Image Classification Method That Uses neither CNNs nor Attention[Paper summury] MLP-mix4 min read·May 19, 2021----
Makoto TAKAMATSUinAIメディカルサービス[論文紹介] MLP-Mixer: An all-MLP Architecture for VisionCNNもAttentionも使わない有用な画像分類手法9 min read·May 19, 2021----