Research map / Vision-language and multimodal / Cross-modal learning
RegBN: Batch Normalization of Multimodal Data with Regularization
MORTEZA GHAHREMANI, Christian Wachinger
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add RegBN to AVVP | AVVP · LLP audio-visual video parsing | event-level event accuracy (%): 48.0 -> 49.3 | Improved |
| add RegBN to AVVP | AVVP · LLP audio-visual video parsing | segment-level event accuracy (%): 55.4 -> 55.2 | No clear effect |
| add RegBN to AVVP | AVVP · LLP audio-visual video parsing | event-level audio accuracy (%): 51.3 -> 52.0 | Improved |
| add RegBN to SMIL | SMIL · MM-IMDb | samples F1 (%): 49.62 -> 54.82 | Improved |
| add PMDN to SMIL | SMIL · MM-IMDb | samples F1 (%): 49.62 -> 49.85 | No clear effect |
| add RegBN to SMIL | SMIL · AV-MNIST | accuracy (%): 98.68 -> 99.11 | Improved |
| add RegBN to MulT | MulT · IEMOCAP word-aligned | Natural accuracy (%): 69.5 -> 73.4 | Improved |
| add RegBN to MulT | MulT · IEMOCAP word-aligned | Happy F1 (%): 84.0 -> 83.0 | Hurt |
| add RegBN to MulT | MulT · CMU-MOSEI word-aligned | Acc2 (%): 80.3 -> 81.1 | Improved |
| add RegBN to MulT | MulT · CMU-MOSEI word-aligned | correlation: 67.1 -> 66.6 | Hurt |
| add RegBN to diagnosis network instead of BN | ADNI network · ADNI diagnosis | test accuracy (%): 48.8 -> 53.0 | Improved |
| add RegBN to MSVT | MSVT · Vision&Touch | test accuracy (%): 86.2 -> 91.5 | Improved |
Closest papers by meaning
- MultiZoo and MultiBench: A Standardized Toolkit for Multimodal Deep Learning (JMLR 2023 · similarity 0.86)
- Mode Normalization (ICLR 2019 · similarity 0.85)
- Training Faster by Separating Modes of Variation in Batch-Normalized Models (TPAMI 2019 · similarity 0.85)
- MultiBench: Multiscale Benchmarks for Multimodal Representation Learning (NeurIPS 2021 · similarity 0.85)
- Filter Response Normalization Layer: Eliminating Batch Dependence in the Training of Deep Neural Networks (CVPR 2020 · similarity 0.85)
- Normalizing the Normalizers: Comparing and Extending Network Normalization Schemes (ICLR 2017 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
