hdong51 commited on
Commit
11ac372
1 Parent(s): 88c9ab8

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -2
README.md CHANGED
@@ -47,6 +47,9 @@ size_categories:
47
 
48
  MultiOOD is the first-of-its-kind benchmark for Multimodal OOD Detection, characterized by diverse dataset sizes and varying modality combinations.
49
 
 
 
 
50
  ## MultiOOD Benchmark
51
  MultiOOD is based on five public action recognition datasets (HMDB51, UCF101, EPIC-Kitchens, HAC, and Kinetics-600).
52
 
@@ -188,8 +191,6 @@ The splits for Multimodal Near-OOD and Far-OOD Benchmarks are provided in https:
188
 
189
  An overview of the proposed framework for Multimodal OOD Detection. We introduce A2D algorithm to encourage enlarging the prediction discrepancy across modalities. Additionally, we propose a novel outlier synthesis algorithm, NP-Mix, designed to explore broader feature spaces, which complements A2D to strengthen the OOD detection performance.
190
 
191
- ## Code
192
- https://github.com/donghao51/MultiOOD
193
 
194
  ## Contact
195
  If you have any questions, please send an email to donghaospurs@gmail.com
 
47
 
48
  MultiOOD is the first-of-its-kind benchmark for Multimodal OOD Detection, characterized by diverse dataset sizes and varying modality combinations.
49
 
50
+ ## Code
51
+ https://github.com/donghao51/MultiOOD
52
+
53
  ## MultiOOD Benchmark
54
  MultiOOD is based on five public action recognition datasets (HMDB51, UCF101, EPIC-Kitchens, HAC, and Kinetics-600).
55
 
 
191
 
192
  An overview of the proposed framework for Multimodal OOD Detection. We introduce A2D algorithm to encourage enlarging the prediction discrepancy across modalities. Additionally, we propose a novel outlier synthesis algorithm, NP-Mix, designed to explore broader feature spaces, which complements A2D to strengthen the OOD detection performance.
193
 
 
 
194
 
195
  ## Contact
196
  If you have any questions, please send an email to donghaospurs@gmail.com