Efficient Indexing of Billion-Scale Datasets of Deep Descriptors

321 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Efficient and Robust Approximate Nearest Neighbor Search Using Hierarchical Navigable Small World Graphs (2016, background): “It would also be interesting to compare Hierarchical NSW on the full 1B SIFT and 1B DEEP datasets [10], [11], [12], [13], [14] and add support for element updates and removal.”
The Faiss Library (2024, unknown): “Faiss was used as a baseline for the challenge and multiple submissions were derived from it.”
Multiscale Quantization for Fast Similarity Search (2017, methodology): “We evaluate the performance of end-to-end trained multiscale quantization (MSQ) on the SIFT1M [20] and DEEP10M [3] datasets, which are often used in benchmarking the performance of nearest neighbor search.”
Face Recognition Vendor Test (FRVT) part 2 :: identification (2019, background): “Further there has been considerable recent research on indexing, exact [13] and approximate nearest neighbor search [1,13] and fast-search [14,16].”
Milvus: A Purpose-Built Vector Data Management System (2021, methodology): “To be reproducible, we use the following two public datasets to evaluate Milvus: SIFT1B [34] and Deep1B [8].”
MINT: Multi-Vector Search Index Tuning (2025, unknown): “To experiment with more than two columns, we generate semi-synthetic datasets by combining single-feature embedding datasets including GloVe25 [36], GloVe50, GloVe100, GloVe200, SIFT1M [37], Yandex Text-to-Image [30], Deep1M [38], and Music [39].”