Pith. sign in

REVIEW 7 cited by

Vision Transformers for Dense Prediction

Not yet reviewed by Pith; the record is open.

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 2103.13413 v1 pith:2OQDFCA7 submitted 2021-03-24 cs.CV

classification cs.CV
keywords densevisiontransformerswhenarchitecturepredictiontransformeravailable
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

We introduce dense vision transformers, an architecture that leverages vision transformers in place of convolutional networks as a backbone for dense prediction tasks. We assemble tokens from various stages of the vision transformer into image-like representations at various resolutions and progressively combine them into full-resolution predictions using a convolutional decoder. The transformer backbone processes representations at a constant and relatively high resolution and has a global receptive field at every stage. These properties allow the dense vision transformer to provide finer-grained and more globally coherent predictions when compared to fully-convolutional networks. Our experiments show that this architecture yields substantial improvements on dense prediction tasks, especially when a large amount of training data is available. For monocular depth estimation, we observe an improvement of up to 28% in relative performance when compared to a state-of-the-art fully-convolutional network. When applied to semantic segmentation, dense vision transformers set a new state of the art on ADE20K with 49.02% mIoU. We further show that the architecture can be fine-tuned on smaller datasets such as NYUv2, KITTI, and Pascal Context where it also sets the new state of the art. Our models are available at https://github.com/intel-isl/DPT.

Discussion (0). Sign in to comment.

Forward citations

Cited by 7 Pith papers

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

  1. MoGe-3: Fine-Detail Monocular Geometry Estimation with Self-Guided Sparse Volumetric Refinement

    cs.CV 2026-07 conditional novelty 7.0 of 10

    Iterative sparse-3D-convolution refinement in a log-depth voxel shell, instead of 2D image-plane refinement, sharply improves fine-detail geometry in monocular point maps and sets state of the art on local fine-detail...

  2. No Pose, No Problem in 4D: Feed-Forward Dynamic Gaussians from Unposed Multi-View Videos

    cs.CV 2026-05 unverdicted novelty 7.0 of 10

    NoPo4D is the first feed-forward system for dynamic 4D Gaussian splatting from unposed multi-view videos, using velocity decomposition supervised by optical flow and a bidirectional motion encoder.

  3. The Multipath Blind Spot: $K$-Agnostic Robust Calibration for Sparse-Anchor Metric Depth from Frozen Foundations

    cs.CV 2026-07 accept novelty 6.5 of 10

    MRAC gates sparse anchors via Theil–Sen + MAD consistency with a frozen foundation's relative depth, repairing multipath outliers that collapse residual-on-CFA and blind VI-Depth while winning 84% of same-backbone cells.

  4. RayOcc: Occlusion-Aware Ray Occupancy Estimation via Gaussian Mixture Intensity

    cs.CV 2026-07 conditional novelty 6.0 of 10

    RayOcc models each camera ray as a non-normalized Gaussian mixture with Poisson-based occupancy probabilities, allowing multiple depth hypotheses per ray and improving Gaussian-initialized 3D occupancy prediction on nuScenes.

  5. Turning Video Models into Generalist Robot Policies

    cs.RO 2026-05 unverdicted novelty 6.0 of 10

    Decouples action-free video world models from embodiment-specific IDMs using Jacobian-based translation to achieve zero-shot cross-embodiment robot policies.

  6. MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer

    cs.CV 2021-10 unverdicted novelty 6.0 of 10

    MobileViT is a lightweight vision transformer that reports 78.4% top-1 accuracy on ImageNet-1k with ~6M parameters, outperforming MobileNetv3 by 3.2% and DeIT by 6.2% at similar size, plus gains on MS-COCO detection.

  7. FF3R: Feedforward Feature 3D Reconstruction from Unconstrained views

    cs.CV 2026-04 unverdicted novelty 5.0 of 10

    FF3R unifies geometric and semantic 3D reconstruction in a single annotation-free feed-forward network trained solely via RGB and feature rendering supervision.

Pith tools