GIM: Learning Generalizable Image Matcher From Internet Videos
Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel pith:USBM4ZIIrecord.jsonopen to challenge →
read the original abstract
Image matching is a fundamental computer vision problem. While learning-based methods achieve state-of-the-art performance on existing benchmarks, they generalize poorly to in-the-wild images. Such methods typically need to train separate models for different scene types and are impractical when the scene type is unknown in advance. One of the underlying problems is the limited scalability of existing data construction pipelines, which limits the diversity of standard image matching datasets. To address this problem, we propose GIM, a self-training framework for learning a single generalizable model based on any image matching architecture using internet videos, an abundant and diverse data source. Given an architecture, GIM first trains it on standard domain-specific datasets and then combines it with complementary matching methods to create dense labels on nearby frames of novel videos. These labels are filtered by robust fitting, and then enhanced by propagating them to distant frames. The final model is trained on propagated data with strong augmentations. We also propose ZEB, the first zero-shot evaluation benchmark for image matching. By mixing data from diverse domains, ZEB can thoroughly assess the cross-domain generalization performance of different methods. Applying GIM consistently improves the zero-shot performance of 3 state-of-the-art image matching architectures; with 50 hours of YouTube videos, the relative zero-shot performance improves by 8.4%-18.1%. GIM also enables generalization to extreme cross-domain data such as Bird Eye View (BEV) images of projected 3D point clouds (Fig. 1(c)). More importantly, our single zero-shot model consistently outperforms domain-specific baselines when evaluated on downstream tasks inherent to their respective domains. The video presentation is available at https://www.youtube.com/watch?v=FU_MJLD8LeY.
This paper has not been read by Pith yet.
Forward citations
Cited by 6 Pith papers
-
MV-Forcing: Long Multi-View Video Generation via 4D-Grounded Spatio-Temporal Self-Forcing
MV-Forcing composes temporal and view-sequential autoregression in a single diffusion model, using a recurrent 3D reconstruction model as a geometric bridge to generate arbitrarily long, multi-view consistent videos.
-
Reshoot-Anything: A Self-Supervised Model for In-the-Wild Video Reshooting
Reshoot-Anything trains a diffusion transformer on pseudo multi-view triplets created by cropping and warping monocular videos to achieve temporally consistent video reshooting with robust camera control on dynamic scenes.
-
ViBA: Implicit Bundle Adjustment with Geometric and Temporal Consistency for Robust Visual Matching
ViBA enables continuous online learning of visual features by embedding implicit differentiable bundle adjustment into a visual odometry pipeline, improving localization accuracy on EuRoC and UMA datasets.
-
RTE-FM-Dehazer: Radiative Transfer Equation Inspired Flow Matching for Real-World Image Dehazing
RTE-FM-Dehazer trains a flow-matching model with an RTE-derived diffusion-absorption regularizer on a new 50k real-haze dataset and reports leading results on five real-world dehazing benchmarks.
-
VLM3: Vision Language Models Are Native 3D Learners
Standard VLMs achieve expert-level 3D performance on depth estimation, pose estimation, and object understanding via three simple techniques without architecture changes or regression losses.
-
AnchorWorld: Embodied Egocentric World Simulation with View-based Evolution Customization
AnchorWorld proposes a simulation framework that adds exogenous viewpoint supervision for full-body grounding and anchor-view text customization for dynamic world evolution in egocentric settings.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.