| Paper Abstract and Keywords |
| Presentation |
2026-08-24 13:45
Improving Open-Vocabulary Semantic Segmentation via Label Clustering
Ryosuke Shimizu, Nagito Saito, Koichi Ito, Takafumi Aoki (Tohoku Univ.) |
| Abstract |
(in Japanese) |
(See Japanese page) |
| (in English) |
Open-Vocabulary Semantic Segmentation (OVSS) is a technique that classifies each pixel in an image based on arbitrary text descriptions. Because OVSS performs inference using semantic similarities between images and language, semantic consistency among class labels is important. However, existing datasets often define semantically similar or overlapping concepts, such as “chair” and “seat” or “building” and “house,” as separate labels. Consequently, even when a model predicts a label that is semantically close to the ground-truth label, the prediction is regarded as a misclassification, which may lead to an underestimation of the model’s inherent recognition performance. To address this issue, this paper proposes a method for improving OVSS accuracy by integrating similar labels and using the resulting labels as text inputs during inference. Specifically, the proposed method clusters class labels based on CLIP text features and concatenates the labels within each cluster into a single text string. Experiments on publicly available datasets demonstrate that the proposed method reduces apparent misclassifications and improves segmentation accuracy. |
| Keyword |
(in Japanese) |
(See Japanese page) |
| (in English) |
Open-Vocabulary Semantic Segmentation / Image Segmentation / CLIP / Clustering / / / / |
| Reference Info. |
ITE Tech. Rep., vol. 50, no. 23, ME2026-56, pp. 5-8, Aug. 2026. |
| Paper # |
ME2026-56 |
| Date of Issue |
2026-08-17 (ME) |
| ISSN |
Online edition: ISSN 2424-1970 |
| Download PDF |
|
| Conference Information |
| Committee |
ME JSPE-IAIP |
| Conference Date |
2026-08-24 - 2026-08-25 |
| Place (in Japanese) |
(See Japanese page) |
| Place (in English) |
Nagoya Kasadera Hotel |
| Topics (in Japanese) |
(See Japanese page) |
| Topics (in English) |
|
| Paper Information |
| Registration To |
ME |
| Conference Code |
2026-08-ME-IAIP |
| Language |
Japanese |
| Title (in Japanese) |
(See Japanese page) |
| Sub Title (in Japanese) |
(See Japanese page) |
| Title (in English) |
Improving Open-Vocabulary Semantic Segmentation via Label Clustering |
| Sub Title (in English) |
|
| Keyword(1) |
Open-Vocabulary Semantic Segmentation |
| Keyword(2) |
Image Segmentation |
| Keyword(3) |
CLIP |
| Keyword(4) |
Clustering |
| Keyword(5) |
|
| Keyword(6) |
|
| Keyword(7) |
|
| Keyword(8) |
|
| 1st Author's Name |
Ryosuke Shimizu |
| 1st Author's Affiliation |
Tohoku University (Tohoku Univ.) |
| 2nd Author's Name |
Nagito Saito |
| 2nd Author's Affiliation |
Tohoku University (Tohoku Univ.) |
| 3rd Author's Name |
Koichi Ito |
| 3rd Author's Affiliation |
Tohoku University (Tohoku Univ.) |
| 4th Author's Name |
Takafumi Aoki |
| 4th Author's Affiliation |
Tohoku University (Tohoku Univ.) |
| 5th Author's Name |
|
| 5th Author's Affiliation |
() |
| 6th Author's Name |
|
| 6th Author's Affiliation |
() |
| 7th Author's Name |
|
| 7th Author's Affiliation |
() |
| 8th Author's Name |
|
| 8th Author's Affiliation |
() |
| 9th Author's Name |
|
| 9th Author's Affiliation |
() |
| 10th Author's Name |
|
| 10th Author's Affiliation |
() |
| 11th Author's Name |
|
| 11th Author's Affiliation |
() |
| 12th Author's Name |
|
| 12th Author's Affiliation |
() |
| 13th Author's Name |
|
| 13th Author's Affiliation |
() |
| 14th Author's Name |
|
| 14th Author's Affiliation |
() |
| 15th Author's Name |
|
| 15th Author's Affiliation |
() |
| 16th Author's Name |
|
| 16th Author's Affiliation |
() |
| 17th Author's Name |
|
| 17th Author's Affiliation |
() |
| 18th Author's Name |
|
| 18th Author's Affiliation |
() |
| 19th Author's Name |
|
| 19th Author's Affiliation |
() |
| 20th Author's Name |
|
| 20th Author's Affiliation |
() |
| 21st Author's Name |
|
| 21st Author's Affiliation |
() |
| 22nd Author's Name |
|
| 22nd Author's Affiliation |
() |
| 23rd Author's Name |
|
| 23rd Author's Affiliation |
() |
| 24th Author's Name |
|
| 24th Author's Affiliation |
() |
| 25th Author's Name |
|
| 25th Author's Affiliation |
() |
| 26th Author's Name |
/ / |
| 26th Author's Affiliation |
()
() |
| 27th Author's Name |
/ / |
| 27th Author's Affiliation |
()
() |
| 28th Author's Name |
/ / |
| 28th Author's Affiliation |
()
() |
| 29th Author's Name |
/ / |
| 29th Author's Affiliation |
()
() |
| 30th Author's Name |
/ / |
| 30th Author's Affiliation |
()
() |
| 31st Author's Name |
/ / |
| 31st Author's Affiliation |
()
() |
| 32nd Author's Name |
/ / |
| 32nd Author's Affiliation |
()
() |
| 33rd Author's Name |
/ / |
| 33rd Author's Affiliation |
()
() |
| 34th Author's Name |
/ / |
| 34th Author's Affiliation |
()
() |
| 35th Author's Name |
/ / |
| 35th Author's Affiliation |
()
() |
| 36th Author's Name |
/ / |
| 36th Author's Affiliation |
()
() |
| Speaker |
Author-1 |
| Date Time |
2026-08-24 13:45:00 |
| Presentation Time |
75 minutes |
| Registration for |
ME |
| Paper # |
ME2026-56 |
| Volume (vol) |
vol.50 |
| Number (no) |
no.23 |
| Page |
pp.5-8 |
| #Pages |
4 |
| Date of Issue |
2026-08-17 (ME) |