Update README with onnx/coreml structure

megaloc/README.md

@@ -0,0 +1,80 @@
+# MegaLoc Quantized Models
+
+Quantized ONNX and CoreML models for [MegaLoc](https://arxiv.org/abs/2502.17237) — SOTA visual place recognition.
+
+## Directory
+
+```
+megaloc/
+├── onnx/         # ONNX models (cross-platform)
+│   ├── megaloc.onnx           # FP32 (873 MB, batch=1)
+│   ├── megaloc_fp16.onnx      # FP16 (437 MB, dynamic batch)
+│   └── megaloc_int8.onnx      # INT8 (223 MB, batch=1)
+└── coreml/       # CoreML models (macOS / Apple Silicon, all multi-batch)
+    ├── megaloc_coreml_fp32.mlpackage   # FP32 (873 MB)
+    ├── megaloc_coreml_fp16.mlpackage   # FP16 (437 MB)
+    └── megaloc_coreml_int8.mlpackage   # INT8 (219 MB)
+```
+
+## Benchmark (MacBook Pro M-series, 518×518 input)
+
+### ONNX (CPU)
+
+| Model | Size | CosSim | B=1 | B=2 | B=4 | B=8 |
+|-------|------|--------|-----|-----|-----|-----|
+| FP32 | 873 MB | 1.000000 | 399ms / 2.5fps | 872ms / 2.3fps | 1514ms / 2.6fps | 2699ms / 3.0fps |
+| FP16 | 437 MB | 0.999989 | 400ms / 2.5fps | 850ms / 2.4fps | 1602ms / 2.5fps | 2876ms / 2.8fps |
+| INT8 | 223 MB | 0.996453 | 475ms / 2.1fps | — | — | — |
+
+FP16 is near-lossless at all batch sizes. INT8 preserves Recall@1 = 100% in retrieval tasks.
+
+### CoreML (Apple Silicon GPU, multi-batch)
+
+| Model | Size | CosSim | B=1 latency | Notes |
+|-------|------|--------|-------------|-------|
+| FP32 | 873 MB | 1.000000 | ~74ms / 13fps | RangeDim(1,16) dynamic batch |
+| FP16 | 437 MB | TBD | TBD | RangeDim(1,16) dynamic batch |
+| INT8 | 219 MB | TBD | TBD | RangeDim(1,16) dynamic batch |
+
+CoreML is ~5× faster than ONNX Runtime (GPU vs CPU). All CoreML models support dynamic batch via `RangeDim(1, 16)`.
+
+## Usage
+
+### ONNX
+
+```python
+import onnxruntime as ort, numpy as np
+from PIL import Image
+
+sess = ort.InferenceSession('megaloc_fp16.onnx', providers=['CPUExecutionProvider'])
+
+img = Image.open('image.jpg').convert('RGB').resize((518, 518), Image.LANCZOS)
+data = np.expand_dims(np.array(img).astype(np.float32).transpose(2,0,1)/255.0, 0)
+
+descriptor = sess.run(None, {'images': data})[0]  # [B, 8448]
+```
+
+### CoreML (macOS only)
+
+```python
+import coremltools as ct, numpy as np
+from PIL import Image
+
+model = ct.models.MLModel('megaloc_coreml_fp16.mlpackage')
+
+img = Image.open('image.jpg').convert('RGB').resize((518, 518), Image.LANCZOS)
+data = np.expand_dims(np.array(img).astype(np.float32).transpose(2,0,1)/255.0, 0)
+
+descriptor = list(model.predict({'images': data}).values())[0]  # [B, 8448]
+```
+
+## Reference
+
+```bibtex
+@inproceedings{Berton_2025_MegaLoc,
+    author    = {Berton, Gabriele and Masone, Carlo},
+    title     = {MegaLoc: One Retrieval to Place Them All},
+    booktitle = {CVPR Workshops},
+    year      = {2025},
+}
+```