Pith. sign in

REVIEW 5 cited by

Improving Multi-Modal Learning with Uni-Modal Teachers

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2106.11059 v1 pith:3HPHUJYN submitted 2021-06-21 cs.LG

Chenzhuang Du , Tingle Li , Yichen Liu , Zixin Wen , Tianyu Hua , Yue Wang , Hang Zhao This is my paper · ORCID
classification cs.LG
keywords multi-modallearningmodalityfusionmethodtaskuni-modalfailure
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Learning multi-modal representations is an essential step towards real-world robotic applications, and various multi-modal fusion models have been developed for this purpose. However, we observe that existing models, whose objectives are mostly based on joint training, often suffer from learning inferior representations of each modality. We name this problem Modality Failure, and hypothesize that the imbalance of modalities and the implicit bias of common objectives in fusion method prevent encoders of each modality from sufficient feature learning. To this end, we propose a new multi-modal learning method, Uni-Modal Teacher, which combines the fusion objective and uni-modal distillation to tackle the modality failure problem. We show that our method not only drastically improves the representation of each modality, but also improves the overall multi-modal task performance. Our method can be effectively generalized to most multi-modal fusion approaches. We achieve more than 3% improvement on the VGGSound audio-visual classification task, as well as improving performance on the NYU depth V2 RGB-D image segmentation task.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Boosting Multimodal Learning via Disentangled Gradient Learning

    cs.CV 2025-07 conditional novelty 6.0 of 10

    Disentangled gradient learning replaces the multimodal gradient to each encoder with a unimodal gradient computed via modality dropout, improving both unimodal and multimodal accuracy across several tasks.

  2. Dynamic Mixture of Curriculum LoRA Experts for Continual Multimodal Instruction Tuning

    cs.CV 2025-06 conditional novelty 6.0 of 10

    D-MoLE dynamically allocates LoRA experts to the most gradient-sensitive layers of a vision-language model for each new task, and routes inputs via per-task autoencoders, reporting near-zero forgetting on a nine-task ...

  3. Improving Multimodal Learning via Imbalanced Learning

    cs.CV 2025-07 conditional novelty 5.0 of 10

    Asymmetric Representation Learning reweights each modality's gradient by the inverse of its prediction variance, improving multimodal accuracy on CREMA-D, Kinetics-Sounds, AVE, MOSI, and UCF101.

  4. Improving Multimodal Learning Balance and Sufficiency through Data Remixing

    cs.LG 2025-06 conditional novelty 5.0 of 10

    Data Remixing improves multimodal learning by decoupling samples into per-modality subsets and training each batch on a single modality, yielding accuracy gains on CREMAD and Kinetic-Sounds.

  5. Decoding Visual Neural Representations by Multimodal with Dynamic Balancing

    cs.CV 2025-09 conditional novelty 4.0 of 10

    A multimodal EEG-image-text contrastive framework with dynamic gradient balancing and stochastic noise improves zero-shot object recognition from EEG on ThingsEEG, raising top-1 accuracy from 13.8% to 15.8%.

Pith tools