Papers › mPLUG: Effective and Efficient Vision-Language Learning by Cross-modal Skip-connections

mPLUG: Effective and Efficient Vision-Language Learning by Cross-modal Skip-connections

24 May 2022arXiv:2205.12005archive 2025-07-28

Chenliang Li, Haiyang Xu, Junfeng Tian, Wei Wang, Ming Yan, Bin Bi, Jiabo Ye, Hehong Chen, Guohai Xu, Zheng Cao, Ji Zhang, Songfang Huang, Fei Huang, Jingren Zhou, Luo Si

Large-scale pretrained foundation models have been an emerging paradigm for building artificial intelligence (AI) systems, which can be quickly adapted to a wide range of downstream tasks. This paper presents mPLUG, a new vision-language foundation model for both cross-modal understanding and generation. Most existing pre-trained models suffer from the problems of low computational efficiency and information asymmetry brought by the long visual sequence in cross-modal alignment. To address these problems, mPLUG introduces an effective and efficient vision-language architecture with novel cross-modal skip-connections, which creates inter-layer shortcuts that skip a certain number of layers for time-consuming full self-attention on the vision side. mPLUG is pre-trained end-to-end on large-scale image-text pairs with both discriminative and generative objectives. It achieves state-of-the-art results on a wide range of vision-language downstream tasks, such as image captioning, image-text retrieval, visual grounding and visual question answering. mPLUG also demonstrates strong zero-shot transferability when directly transferred to multiple video-language tasks.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

alibaba/AliceMind officialpytorch report
x-plug/mplug mentioned on GitHubpytorch report
modelscope/modelscope pytorchApache-2.0 report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Computational EfficiencyImage CaptioningImage-text RetrievalQuestion AnsweringRetrievalText RetrievalVideo-Text RetrievalVisual GroundingVisual Question AnsweringVisual Question Answering (VQA)cross-modal alignment

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Captioning COCO Captions mPLUG BLEU-4 46.5 #1 of 41 Archive leaderboard report
Image Captioning COCO Captions mPLUG CIDER 155.1 #1 of 41 Archive leaderboard report
Image Captioning COCO Captions mPLUG METEOR 32.0 #1 of 41 Archive leaderboard report
Image Captioning COCO Captions mPLUG SPICE 26.0 #1 of 41 Archive leaderboard report
Visual Question Answering (VQA) VQA v2 test-dev mPLUG (Huge) Accuracy 82.43 #5 of 56 Archive leaderboard report
Visual Question Answering (VQA) VQA v2 test-std mPLUG-Huge number 69.82 #2 of 38 Archive leaderboard report
Visual Question Answering (VQA) VQA v2 test-std mPLUG-Huge other 77.02 #2 of 38 Archive leaderboard report
Visual Question Answering (VQA) VQA v2 test-std mPLUG-Huge overall 83.62 #2 of 38 Archive leaderboard report
Visual Question Answering (VQA) VQA v2 test-std mPLUG-Huge yes/no 94.83 #2 of 38 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections