Reliev3R: Relieving Feed-forward Reconstruction from Multi-View Geometric Annotations

2026-04-01 · Youyu Chen, Junjun Jiang, Yueru Luo, Kui Jiang, Xianming Liu, Xu Yan, Dave Zhenyu Chen arxiv

With recent advances, Feed-forward Reconstruction Models (FFRMs) have demonstrated great potential in reconstruction quality and adaptiveness to multiple downstream tasks. However, the excessive reliance on multi-view geometric annotations, e.g. 3D point maps and camera poses, makes the fully-supervised training scheme of FFRMs difficult to scale up. In this paper, we propose Reliev3R, a weakly-supervised paradigm for training FFRMs from scratch without cost-prohibitive multi-view geometric annotations. Relieving the reliance on geometric sensory data and compute-exhaustive structure-from-motion preprocessing, our method draws 3D knowledge directly from monocular relative depths and image sparse correspondences given by zero-shot predictions of pretrained models. At the core of Reliev3R, we design an ambiguity-aware relative depth loss and a trigonometry-based reprojection loss to facilitate supervision for multi-view geometric consistency. Training from scratch with the less data, Reliev3R catches up with its fully-supervised sibling models, taking a step towards low-cost 3D reconstruction supervisions and scalable FFRMs.

📄 PDF Abstract BibTeX arXiv:2604.00548

Code (0)

등록된 구현이 없습니다.

Tasks

3D Reconstruction

Relieving the Computational Bottleneck: Joint Inference for Event Extraction with High-Dimensional Features

2014-10-01 · EMNLP 2014 10 · Deepak Venugopal, Chen Chen, Vibhav Gogate, Vincent Ng

Event Extraction

Meta-Learning Probabilistic Inference For Prediction

2018-05-24 · ICLR 2019 5 · Jonathan Gordon, John Bronskill, Matthias Bauer, Sebastian Nowozin 외

This paper introduces a new framework for data efficient and versatile learning. Specifically: 1) We develop ML-PIP, a general framework for Meta-Learning approximate Probabilistic Inference for Prediction. ML-PIP extend…

Few-Shot Image Classification Few-Shot Learning Meta-Learning Prediction

Memory-Efficient Differentiable Transformer Architecture Search

2021-05-31 · Findings (ACL) 2021 8 · Yuekai Zhao, Li Dong, Yelong Shen, Zhihua Zhang 외

Differentiable architecture search (DARTS) is successfully applied in many vision tasks. However, directly using DARTS for Transformers is memory-intensive, which renders the search process infeasible. To this end, we pr…

Measuring and Relieving the Over-smoothing Problem for Graph Neural Networks from the Topological View

2019-09-07 · Deli Chen, Yankai Lin, Wei Li, Peng Li 외

Graph Neural Networks (GNNs) have achieved promising performance on a wide range of graph-based tasks. Despite their success, one severe limitation of GNNs is the over-smoothing issue (indistinguishable representations o…

Node Classification

Segregated Temporal Assembly Recurrent Networks for Weakly Supervised Multiple Action Detection

2018-11-19 · Yunlu Xu, Chengwei Zhang, Zhanzhan Cheng, Jianwen Xie 외

This paper proposes a segregated temporal assembly recurrent (STAR) network for weakly-supervised multiple action detection. The model learns from untrimmed videos with only supervision of video-level labels and makes pr…

Action Detection Multiple Action Detection