rii 0.1.0

Fast and memory-efficient ANN with a subset-search functionality

Rii

Reconfigurable Inverted Index (Rii): fast and memory efficient approximate nearest neighbor search method with a subset-search functionality.

Reference:

• Y. Matsui, R. Hinami, and S. Satoh, "Reconfigurable Inverted Index", ACM Multimedia 2018 (oral). [paper] [supplementary] [project]

Summary of features

• Fast and memory efficient ANN. Can handle billion-scale data on memory at once. The search is less than 10 ms.
• Can run the search over a subset of the whole database
• Remain fast even after a large number of vectors are newly added (i.e., the data structure can be reconfigured)

Installing

You can install the package via pip. This library works with Python 3.5+ on linux.

pip install rii

Documentation

• Tutorial
• API

Example

Basic ANN

import rii
import numpy as np

N, D = 10000, 128
X = np.random.random((N, D)).astype(np.float32)  # 10,000 128-dim vectors
q = np.random.random((D,)).astype(np.float32)  # a 128-dim vector

# Instantiate with M=32 sub-spaces
e = rii.Rii(M=32)

# Train with the top 1000 vectors
e.fit(vecs=X[:1000])

# Add vectors
e.add_reconfigure(vecs=X)

# Search
ids, dists = e.query(q=q, topk=5)
print(ids, dists)  # e.g., [1965  951 1079] [11.77537251 13.04397392 13.06065941]

Subset search

# The search can be conducted over a subset of the database
target_ids = np.array([85, 132, 236, 551, 694, 728, 992, 1234])  # Specified by IDs
ids, dists = e.query(q=q, topk=3, target_ids=target_ids)
print(ids, dists)  # e.g., [728  85 132] [14.80522156 15.92787838 16.28690338]

Reconfigure

# Add new vectors
X2 = np.random.random((1000, D)).astype(np.float32)
e.add(vecs=X2)  # Now N is 11000
e.query(q=q)  # Ok. (0.12 msec / query)

# However, if you add quite a lot of vectors, the search might become slower
# because the data structure has been optimized for the initial item size (N=10000)
X3 = np.random.random((1000000, D)).astype(np.float32) 
e.add(vecs=X3)  # A lot. Now N is 1011000
e.query(q=q)  # Slower (0.96 msec/query)

# In such case, run the reconfigure function. That updates the data structure
e.reconfigure()
e.query(q=q)  # Ok. (0.21 msec / query)

Author

• Yusuke Matsui

Citation

@inproceedings{rii,
  author = {Yusuke Matsui and Ryota Hinami and Shin'ichi Satoh},
  title = {Reconfigurable Inverted Index},
  booktitle = {ACM International Conference on Multimedia (ACMMM)},
  year = {2018},
}