Skip to content
ModelListedExternal
Download

Geospot Base

A geolocation model built on SigLIP2-so400m (512px) that predicts GPS coordinates from images.

Sourcehuggingface.co/sdan/geospot-base/tree/32f164c81d62bc880d7f8c4fc3092a0d36eac1a2
KindOtherVersionRevision 32f164cUnratedgrade

Model source

Kind
Other
File size
861 MB
Version
vRevision 32f164c
Updated
Nov 5, 2026
Downloads
27
Source
Hugging Face

Source description

GeoSpot Base

A geolocation model built on SigLIP2-so400m (512px) that predicts GPS coordinates from images.

Model Details

  • Backbone: google/siglip2-so400m-patch16-512 (frozen)
  • Image Resolution: 512x512
  • Embedding Dim: 512
  • Training Steps: 206k
  • Training Data: ~10.6M streetview images

Architecture

GeoCLIP-style contrastive learning between:

  • Image Encoder: SigLIP2 vision tower + MLP projection (1152 → 512)
  • Location Encoder: Multi-scale RFF encoding with learnable capsules

Usage

from geoclip.model.GeoCLIP import GeoCLIP
import torch

model = GeoCLIP(from_pretrained=False, encoder_name="siglip2")
state_dict = torch.load("model.safetensors")
model.load_state_dict(state_dict)

# Predict location from image
top_gps, top_probs = model.predict("image.jpg", top_k=5)

Sources

1 source
Hugging FacePrimaryVersion 32f164c81d62bc880d7f8c4fc3092a0d36eac1a2
huggingface.co/sdan/geospot-base/tree/32f164c81d62bc880d7f8c4fc3092a0d36eac1a2

Verified Nov 5

Download files

Repository files

File inventory only; completeness is unverified.

Model artifacts

1 artifact

model.safetensors

safetensors · 861 MB · SHA-256 d859b188a68a…00ac · Hugging Face

Download file

Source excerpts

3 excerpts
ReadmeAug 26huggingface.co/sdan/geospot-base

--- license: apache-2.0 tags: - geolocation - vision - siglip - clip - geoclip datasets: - osv5m pipeline_tag: image-feature-extraction --- # GeoSpot Base A geolocation model built on SigLIP2-so400m (512px) that predicts GPS coordinates from images. ## Model Details - **Backbone**: google/siglip2-so400m-patch16-512 (frozen) - **Image Resolution**: 512x512 - **Embedding Dim**: 512 - **Training Steps**: 206k - **Training Data**: ~10.6M streetview images ## Architecture GeoCLIP-style contrastive learning between: - **Image Encoder**: SigLIP2 vision tower + MLP projection (1152 → 512) - **Location Encoder**: Multi-scale RFF encoding with learnable capsules ## Usage ```python from geoclip.model.GeoCLIP import GeoCLIP import torch model = GeoCLIP(from_pretrained=False, encoder_name="siglip2") state_dict = torch.load("model.safetensors") model.load_state_dict(state_dict) # Predict location from image top_gps, top_probs = model.predict("image.jpg", top_k=5) ```

NoteNov 5huggingface.co/sdan/geospot-base

Source context: 27 downloads · 2 likes · Pipeline image-feature-extraction · Repo sdan/geospot-base

NoteAug 26huggingface.co/sdan/geospot-base

Source context: 16 downloads · 2 likes · Pipeline image-feature-extraction · Repo sdan/geospot-base