AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

4M: Massively Multimodal Masked Modeling

David Mizrahi, Roman Bachmann, Oguzhan Fatih Kar, Teresa Yeo, Mingfei Gao, Afshin Dehghan, Amir Zamir

NeurIPS 2023 · Vision-language and multimodal › Human and embodied modeling · approach: Vision-language pretraining · 9 citations (OpenAlex)

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 changedWhereMeasuredEffect
pre-train with all target modalities instead of RGB only (RGB inputs) (ablation)4M · average transfer lossavg loss: 5.49 -> 5.29Improved
pre-train with all inputs and all targets instead of RGB inputs (avg loss) (ablation)4M · average transfer lossavg loss: 5.29 -> 5.36Mixed
use multimodal inputs (All) vs RGB-only inputs on X->Y transfers (ablation)4M · X->Y transfer losscross-entropy: 6.91 -> 6.75Improved
use multimodal inputs vs RGB-only on RGB->X transfers (ablation)4M · RGB->X transfer losscross-entropy: 4.74 -> 5.07Hurt
add captions as the only target (ablation)4M · average transfer lossavg loss: 5.49 -> 6.03Hurt
multi-target CLIP pre-training (4M-B) vs RGB -> CLIP only (ablation)4M-B · COCOAP box: 46.6 -> 49.7Improved
multi-modal multitask pre-training (4M-B) vs RGB -> RGB only (ablation)4M-B · NYU depthdelta1 accuracy: 80.4 -> 92.0Improved
4M-B vs MAE B4M-B · COCOAP box: 48.3 -> 49.7Improved
4M-B vs DeiT III B4M-B · ImageNet-1Ktop-1 accuracy: 85.4 -> 84.5Hurt
4M-L vs MAE L4M-L · NYU depthdelta1 accuracy: 93.6 -> 94.4Improved
increase Dirichlet alpha (uniform simplex sampling) (ablation)4M · avg transfer lossavg lossImproved
decode only a small random subset of targets (target masking) (ablation)4M · avg transfer lossavg lossImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy