pith:LX2ZLAUE
REALM: An RGB- and Event-Aligned Latent Manifold for Cross-Modal Perception
Event data can be projected into pretrained RGB latent spaces using low-rank adaptation to enable zero-shot application of frozen image decoders to raw event streams.
arxiv:2605.00271 v3 · 2026-04-30 · cs.CV · cs.AI · cs.RO
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{LX2ZLAUENEMKJO76JZAJGZTCEE}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
REALM enables the direct, zero-shot application of complex, frozen image-trained decoders, such as MASt3R, to raw event data. We demonstrate state-of-the-art performance in wide-baseline feature matching, significantly outperforming specialized architectures.
That low-rank adaptation of frozen RGB backbones is sufficient to bridge the modality gap and unlock geometric and semantic priors for asynchronous event streams without task-specific training or substantial performance loss.
REALM aligns event streams with RGB ViT latent spaces via LoRA, enabling zero-shot transfer of image-trained linear heads and complex decoders like MASt3R to raw event data with reported SOTA wide-baseline matching performance.
Receipt and verification
| First computed | 2026-07-02T01:17:31.889775Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
5df59582846918a4bbfe4e40936662212208c14917e55c71c788c76b0bd3985e
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/LX2ZLAUENEMKJO76JZAJGZTCEE \
| jq -c '.canonical_record' \
| python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: 5df59582846918a4bbfe4e40936662212208c14917e55c71c788c76b0bd3985e
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "0eb613baf49d2f00ea973a5b1a8fadf322172b6ecd3e808e9a27cc0f643b939d",
"cross_cats_sorted": [
"cs.AI",
"cs.RO"
],
"license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
"primary_cat": "cs.CV",
"submitted_at": "2026-04-30T22:14:36Z",
"title_canon_sha256": "5a62fa8feda0a0bd9674b6cbd3b19ced6f16d86b8afd15d99e7a1e3a41152ea0"
},
"schema_version": "1.0",
"source": {
"id": "2605.00271",
"kind": "arxiv",
"version": 3
}
}