Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
一般物体検出とLSTMを用いた画像に基づく屋内位置推定 - IPSJ UBI82
Search
Aokiti
May 12, 2024
0
390
一般物体検出とLSTMを用いた画像に基づく屋内位置推定 - IPSJ UBI82
http://id.nii.ac.jp/1001/00233750/
Aokiti
May 12, 2024
Tweet
Share
More Decks by Aokiti
See All by Aokiti
[d-hacks Docker講座] Dockerで動かすローカルLLM入門
sakusaku3939
0
18
[論文輪読会] A survey of model compression strategies for object detection
sakusaku3939
0
7
[論文輪読会] ViT-1.58b
sakusaku3939
0
91
d-hacks PyTorchモデル実装会 2024f
sakusaku3939
0
43
[論文輪読会] Binarized Neural Networks
sakusaku3939
0
41
MoodTune 東京AI祭ハッカソン決勝
sakusaku3939
0
510
d-hacks PyTorch実装会 2023f
sakusaku3939
0
25
[DL勉強会] 第5章 ディープラーニングを活用したアプリケーション 後半
sakusaku3939
0
16
Presc - 高校研究発表会
sakusaku3939
0
1.6k
Featured
See All Featured
Bash Introduction
62gerente
614
210k
The Straight Up "How To Draw Better" Workshop
denniskardys
236
140k
Dealing with People You Can't Stand - Big Design 2015
cassininazir
367
26k
Optimising Largest Contentful Paint
csswizardry
37
3.4k
Producing Creativity
orderedlist
PRO
347
40k
Become a Pro
speakerdeck
PRO
29
5.5k
A Modern Web Designer's Workflow
chriscoyier
695
190k
Why Our Code Smells
bkeepers
PRO
338
57k
Connecting the Dots Between Site Speed, User Experience & Your Business [WebExpo 2025]
tammyeverts
8
480
Let's Do A Bunch of Simple Stuff to Make Websites Faster
chriscoyier
507
140k
JavaScript: Past, Present, and Future - NDC Porto 2020
reverentgeek
50
5.5k
Java REST API Framework Comparison - PWX 2021
mraible
33
8.8k
Transcript
1
2
3
4
5
6
[7] F. Walch, C. Hazirbas, L. Leal-Taixé, T. Sattler, S.
Hilsenbeck, D. Cremers 7 https://doi.org/10.1109/ICCV.2017.75
8 [8] S Nilwong, D Hossain, S Kaneko, G Capi
https://doi.org/10.3390/machines7020025
9
10 CNN(GoogLeNetモデル) LSTM(次元を削減) CNN CNN CNN LSTM(1つの特徴量に変換)
11 CNN CNN CNN LSTM(1つの特徴量に変換)
12
13
14
15
16 𝑀𝐴𝐸 = 1 𝑛 𝑖=1 𝑛 |ෝ 𝑥𝑖
− 𝑥𝑖 + | ෝ 𝑦𝑖 − 𝑦𝑖 |)
17 CNN (4層) CNN(4層) CNN(4層) LSTM(1つの特徴量に変換) CNN(4層) CNN(GoogLeNet) CNN(GoogLeNet) LSTM(次元を削減)
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34