The Replica Dataset: A Digital Replica of Indoor Spaces

By Julian Straub, Thomas Whelan, Lingni Ma, Yufan Chen, Erik Wijmans, Simon Green, Jakob J. Engel, Raul Mur-Artal, Carl Yuheng Ren, Shobhit Verma and 20 more. 2019

1,356 citations recorded by the source corpus.

We introduce Replica, a dataset of 18 highly photo-realistic 3D indoor scene reconstructions at room and building scale. Each scene consists of a dense mesh, high-resolution high-dynamic-range (HDR) textures, per-primitive semantic class and instance information, and planar mirror and glass reflectors. The goal of Replica is to enable machine learning (ML) research that relies on visually, geometrically, and semantically realistic generative models of the world - for instance, egocentric computer vision, semantic segmentation in 2D and 3D, geometric inference, and the development of embodied agents (virtual robots) performing navigation, instruction following, and question answering. Due to the high level of realism of the renderings from Replica, there is hope that ML systems trained on Replica may transfer directly to real world image and video data. Together with the data, we are releasing a minimal C++ SDK as a starting point for working with the Replica dataset. In addition, Replica is `Habitat-compatible', i.e. can be natively used with AI Habitat for training and testing embodied agents.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, Geology, Evaluation & Benchmarks, 3D Vision & Pose

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

In Geology

In Evaluation & Benchmarks

In 3D Vision & Pose

Citations in the corpus

Habitat 2.0: Training Home Assistants to Rearrange their Habitat (2021, background): “ReplicaCAD is an artist-created, fully-interactive recreation of ‘FRL-apartment’ spaces from the Replica dataset [2].”
Ego4D: Around the World in 3,000 Hours of Egocentric Video (2021, background): “[209] Julian Straub, Thomas Whelan, Lingni Ma, Yufan Chen, Erik Wijmans, Simon Green, Jakob J Engel, Raul Mur-Artal, Carl Ren, Shobhit Verma, et al.”
Hypersim: A Photorealistic Synthetic Dataset for Holistic Indoor Scene Understanding (2020, methodology): “This level of clutter and annotation detail compares favorably to existing indoor scene datasets (e.g., ScanNet [19]: 14 objects per scene; Replica [72]: 84 objects per scene).”
A Survey of Embodied AI: From Simulators to Research Tasks (2021, methodology): “EQA [102] dataset is based on House3D, a subset of recent popular SUNCG dataset with synthesis rooms and layouts which is similar to the Replica dataset [108].”
Habitat: A Platform for Embodied AI Research (2019, background): “the top row, and a Replica [25] environment in the bottom row.”
SAPIEN: A SimulAted Part-Based Interactive ENvironment (2020, background): “Navigation environments typically use datasets providing real-world RGB-D scans [56, 6, 46], and/or high-quality synthetic scenes [44].”
MSeg: A Composite Dataset for Multi-Domain Semantic Segmentation (2020, background): “REPLICA DATASET [66] Indoor (Synthetic) N/A Real data is not available (images are synthetically rendered).”
Occupancy Anticipation for Efficient Exploration and Navigation (2020, background): “Visual navigation has seen substantial progress in the past few years, fueled by large-scale datasets and photo-realistic 3D environments [4,9,74,69], simulators [74,34,3,38], and public benchmarks [13,3,38].”
SynSin: End-to-End View Synthesis From a Single Image (2019, methodology): “Matterport [4] RealEstate10K [74] Replica [58]”
OpenMask3D: Open-Vocabulary 3D Instance Segmentation (2023, methodology): “We conduct our experiments using the ScanNet200 [57] and Replica [61] datasets.”
NICE-SLAM: Neural Implicit Scalable Encoding for SLAM (2021, methodology): “Geometry optimization on a single depth image on Replica [44] with different architectures.”
NeRF: Neural Radiance Field in 3D Vision, A Comprehensive Review (2022, neutral): “According to the GLO authors, this method can be thought of as a Discriminator-less GAN.”
Co-SLAM: Joint Coordinate and Sparse Parametric Encodings for Neural Real-Time SLAM (2023, methodology): “Following iMAP and NICESLAM, we quantitatively evaluate the reconstruction quality on 8 synthetic scenes from Replica [24].”
ConceptFusion: Open-set Multimodal 3D Mapping (2023, methodology): “The benchmark comprises 20 indoor (apartment-scale) scenes from ScanNet [59, 60], Replica [61], and self-captured sequences; 5 outdoor (urban driving) scenes; 20 indoor (tabletop) scenes with common household products (UnCoCo); and 5 synthetic scenes from the ICL [62] and iTHOR benchmarks [63].”
GS-SLAM: Dense Visual SLAM with 3D Gaussian Splatting (2023, methodology): “Rendering Performance on Replica [28].”
Depth Anything 3: Recovering the Visual Space from Any Views (2025, neutral): “…[98], IRS [93], vKITTI2 [9], BlendedMVS [115], SPRING [56], MVSSynth [38], UnrealStereo4K [123], GTA-SfM [92], TauAgent [27], KenBurns [60], MatrixCity [50], EDEN [47], ReplicaGSO [82], UrbanSyn [32], PointOdyssey [127], Structured3D [125], Objaverse [17], Trel-lis [104], and OmniObject [102].”
ABO: Dataset and Benchmarks for Real-World 3D Object Understanding (2021, background): “There also exist many high quality datasets specifically made for 3D scene understanding [48,50,52,58] rather than individual objects.”
Orbit: A Unified Simulation Framework for Interactive Robot Learning Environments (2023, methodology): “The world can be designed procedurally (script-based), via scanned meshes [28], or interactively through the game-based GUI of Isaac Sim, or a combination of them.”
How to Train PointGoal Navigation Agents on a (Sample and Compute) Budget (2020, background): “Galvanized by fast simulation platforms [15, 20, 21, 29], large rich 3D datasets [4, 25, 30], and the success of deep reinforcement learning [18, 24, 26], training virtual robots (embodied agents) in simulation has garnered considerable interest in recent years.”
SoundSpaces: Audio-Visual Navigation in 3D Environments (2019, methodology): “Replica [83] is a dataset of 18 apartment, hotel, office, and room scenes with 3D meshes.”