Segment-anything model with Microscopy Image Browser

  Рет қаралды 508

Ilya Belevich

Ilya Belevich

Күн бұрын

Demonstration on how to install and use segment-anything model from Meta AI (segment-anything.com) with Microscopy Image Browser (mib.helsinki.fi)
At the current stage there are 7 trained implementations available:
vit_h, vit_l, vit_b - default models coming from the original Meta-AI research (*)
vit_h_em, vit_b_em - fine-tuned models for electron microscopy (**)
vit_h_lm, vit_b_lm - fine-tuned models for light microscopy(**)
References:
* Kirillov A, Mintun E, Ravi N, Mao H, Rolland C, Gustafson L, Xiao T, Whitehead S, Berg AC, Lo W-Y, Dollár P, Girshick R. Segment Anything, arXiv:2304.02643
arxiv.org/abs/2304.02643
** Archit A, Nair S, Khalid N, Hilt P, Rajashekar V, Freitag M, Gupta S, Dengel A, Ahmed S, Pape C. Segment Anything for Microscopy. bioRxiv, doi.org/10.1101/2023.08.21.55...
00:00 Introduction to Segment-anything model
02:18 Requirements of SAM in MIB
03:29 Installation of Miniconda
06:22 Installation of SAM files
07:13 Creating Python environment
11:33 Setting up SAM and Python in MIB
15:34 How to use SAM in MIB, missing driver
17:04 SAM in the Interactive mode
18:20 Add, Replace, Remove modes
22:09 Add, +next material mode
22:57 SAM in the Landmarks mode
25:12 SAM in the Automatic mode

Пікірлер: 6
MIB: Tutorial on 2D patch-wise segmentation using deep learning
32:24
Amazing weight loss transformation !! 😱😱
00:24
Tibo InShape
Рет қаралды 65 МЛН
MIB Brief: updated 3D volume and surface viewer
9:26
Ilya Belevich
Рет қаралды 271
Building Footprint extraction in QGIS.
5:28
Nexxis Tutorials
Рет қаралды 94
DeepMIB: features and updates in MIB 2.80
31:25
Ilya Belevich
Рет қаралды 772
MCcalc part5: Processing with MCcalc
6:18
Ilya Belevich
Рет қаралды 126
Implementing Segment Anything Model in the browser
0:58
Ankit Sanghvi
Рет қаралды 38
MIB Brief: Kymograph
3:32
Ilya Belevich
Рет қаралды 159
تجربة أغرب توصيلة شحن ضد القطع تماما
0:56
صدام العزي
Рет қаралды 63 МЛН
АЙФОН 20 С ФУНКЦИЕЙ ВИДЕНИЯ ОГНЯ
0:59
КиноХост
Рет қаралды 1,2 МЛН
low battery 🪫
0:10
dednahype
Рет қаралды 1,7 МЛН