yali30/findingdory-habitat

FindingDory: A Benchmark to Evaluate Memory in Embodied Agents

Karmesh Yadav*, Yusuf Ali*, Gunshi Gupta, Yarin Gal, Zsolt Kira

Current vision-language models (VLMs) struggle with long-term memory in embodied tasks. To address this, we introduce FindingDory, a benchmark in Habitat that evaluates memory-based reasoning across 60 long-horizon tasks.

In this repo, we release the FindingDory Habitat Dataset. The episode dataset is used to run online evaluations with the Habitat simulator using VLM-based navigation agents. Each episode involves a robot performing multiple pick-place object interactions. The agent then needs to execute actions in the simulator complete the tasks specified in the FindingDory benchmark.

Dataset Structure

This repository contains data and models for the dnb_release package, structured into two main components: findingdory and imgnav.

---

findingdory

File/Folder	Description
train/	Training split of the habitat dataset.
├── episodes.json.gz	Encoded JSON file containing episode metadata for training episodes.
├── transformations.npy	NumPy array of transformation matrices (of each rigid object in simulator).
└── viewpoints.npy	NumPy array of viewpoints (camera poses of all objects/receptacles).
val/	Validation split of the habitat dataset.
├── episodes.json.gz	Encoded JSON file containing episode metadata for validation episodes.
├── transformations.npy	NumPy array of transformation matrices (of each rigid object in simulator).
└── viewpoints.npy	NumPy array of viewpoints (camera poses of all objects/receptacles).

---

findingdory_imagenav

File/Folder	Description
dataset/	Image navigation dataset (split into train and val).
├── train/	Training data for the image navigation task.
└── val/	Validation data for the image navigation task.
policy_ckpt/	Policy checkpoint files (trained navigation policies).
└── ckpt.17.pth	PyTorch checkpoint file containing trained policy weights.
pretrained_vis_enc/	Pretrained visual encoder weights.
└── ckpt.16.pth	PyTorch checkpoint file for the pretrained visual encoder used in imagenav training.

📄 Citation

@article{yadav2025findingdory,
  title     = {FindingDory: A Benchmark to Evaluate Memory in Embodied Agents},
  author    = {Yadav, Karmesh and Ali, Yusuf and Gupta, Gunshi and Gal, Yarin and Kira, Zsolt},
  journal   = {arXiv preprint arXiv:2506.15635},
  year      = {2025}
}