Pith. sign in

REVIEW 3 major objections 5 minor 31 references

Dynamic Exploration Graph: A Novel Approach for Efficient Nearest Neighbor Search in Evolving Multimedia Datasets

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read DEG supports instant vertex deletion with guaranteed connectivity and runs at static-index speed.

desk verdict A practically useful dynamic ANNS graph with immediate deletion, held back by an unproved connectivity guarantee in the repair step. read the letter →

arxiv 2607.27640 v1 pith:75FQWMKB submitted 2026-07-30 cs.IR

Nico Hezel , Kai Uwe Barthel , Bruno Schilling , Konstantin Schall , Klaus Jung This is my paper · ORCID
classification cs.IR
keywords approximatenearestneighborsearchdynamicgraphindexeven-regularconnectivityvertexdeletionstreamingdataonlinemultimediaretrieval
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper introduces the Dynamic Exploration Graph (DEG), a graph-based index for approximate nearest neighbor search that supports immediate insertion and deletion of data points. It claims DEG keeps search speed and recall comparable to the best static graph methods while also handling streaming and online datasets, something most graph indexes cannot do without marking deleted vertices as tombstones. Two mechanisms carry the claim: a deletion routine that reconnects the graph after removing a vertex, and a distribution-agnostic expansion routine that attaches new vertices without a tuned hyperparameter. If correct, DEG would remove a long-standing bottleneck for multimedia retrieval over evolving data.

What carries the argument

The central object is an even-regular undirected graph (degree d, with d at least 4) maintained as an Eulerian graph, so every edge lies on a cycle and the graph is 2-edge-connected. That property is what makes deletion feasible: after a vertex is removed, each of its neighbors still has an alternative path through the graph. The load-bearing mechanism is the restoration routine (Algorithm 2), which forms induced subgraphs from the deleted vertex's neighbors, merges them via simultaneous BFS until d/2−1 components remain, then adds circular edges and repairs degrees so the graph stays regular and connected. The expansion routine (Algorithm 3) uses a range search to find candidate neighbors, applies an MRNG check to prefer edges that will help navigation, and feeds any vertex that lost an edge back into the same restoration routine.

What would settle it

Build a small even-regular graph such as DEG_4 on random points, delete a vertex whose former neighbors lie in separate dense clusters, and run Algorithm 2; if any pair of remaining vertices is disconnected or any vertex ends with degree different from d, the guarantee fails.

Watch

Extended reading notes

Core claim

The central claim is that a graph index can be fully dynamic—both insertions and deletions—without sacrificing the speed and connectivity that make graph-based approximate nearest neighbor search work. DEG is an even-regular undirected graph built on the continuously refining Exploration Graph. To delete a vertex, it isolates the vertex and runs a restoration procedure (Algorithm 2) that merges the remaining components via breadth-first search and adds circular edges among the deleted vertex's former neighbors, restoring both connectivity and degree regularity. To insert a vertex, it attaches to the closest candidates from a range search, occasionally stealing an edge from a saturated neighbor, and then reuses the same restoration routine. In experiments on SIFT1M, Deep1M, and GloVe, the paper reports DEG builds faster than HNSW, DiskANN, and SWINN, and achieves higher queries per second at recall above 95% in static, streaming, and online scenarios.

Load-bearing premise

The connectivity guarantee rests on an unproved claim that the restoration procedure in Algorithm 2 always reconnects every component left by a deleted vertex; the paper offers no invariant or proof, only a special-case loop.

Editorial extensions

If this is right

  • DEG can be used as a drop-in index for collections that change over time, since it removes vertices immediately instead of marking them deleted, avoiding the slowdown that tombstoning causes in HNSW-style indexes.
  • Construction and deletion costs grow logarithmically with index size, suggesting the approach scales beyond the million-point datasets tested here.
  • Because the same graph structure performs at parity with static state-of-the-art graphs in the static scenario, a single index design can serve static, streaming, and online workloads without reconfiguration.
  • The distribution-agnostic expansion removes a hyperparameter that would otherwise need retuning when the data distribution drifts.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the connectivity guarantee holds, the restoration routine could be adapted to other even-regular graph indexes, providing a general recipe for immediate deletion in navigable small-world graphs.
  • The BFS-based subgraph merging may implicitly rely on local clustering; a testable extension would be measuring whether deletion cost worsens on datasets with high local intrinsic dimension where neighborhoods overlap less.
  • Since the paper compares against tombstoning and periodic consolidation, a natural follow-up is whether DEG's immediate deletion remains competitive under burst deletions where many vertices are removed at once.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes the Dynamic Exploration Graph (DEG), an extension of the authors' prior continuously refining Exploration Graph (crEG), for approximate nearest neighbor search in dynamic datasets. DEG is designed to support both insertions and deletions while maintaining an even-regular undirected graph with guaranteed connectivity. Two claimed innovations are a vertex deletion algorithm that repairs the graph immediately without tombstones, and a data distribution-agnostic graph expansion method. The paper presents pseudocode for reduction, restoration, and extension (Algorithms 1-4), and reports experiments on SIFT1M, Deep1M, and GloVe in static, online, and streaming scenarios, comparing DEG against HNSW, DiskANN, and SWINN. The reported results show DEG with lower construction times and higher search efficiency in most settings.

Significance. If the connectivity and regularity guarantees for the deletion algorithm are correct, DEG would be a practically valuable fully dynamic graph-based ANNS index, avoiding the tombstone overhead of HNSW and the delayed consolidation of FreshDiskANN. The paper includes an open-source implementation, which is a strength for reproducibility. The comparison against external baselines on three standard datasets is a useful empirical contribution. However, the paper's central theoretical claim—that Algorithm 2 always restores connectivity and even regularity after arbitrary deletions—is asserted without proof. Since this guarantee is the basis for the paper's main contribution and for the claim that deletions leave no stale vertices, the manuscript is not yet ready for acceptance. The empirical evaluation also lacks any measure of run-to-run variability, which weakens the strength of the performance claims.

major comments (3)
  1. [Section 5.1, Algorithm 2] The headline claim that the deletion mechanism 'guarantees graph connectivity' (Abstract and Section 5) is not established. No invariant or proof is given that RestoreGraph terminates with a connected, even-regular graph after removing an arbitrary vertex. In particular, Step 5 (lines 30-39) assumes that for every remaining vertex v_A there exists a non-neighbor v_B in its 2-hop neighborhood, and that for each v_C there exists v_D in N(G,v_B) satisfying the stated non-adjacency conditions. These existential requirements are asserted but not derived from the graph's even-regularity, Eulerian property, or edge-count parity. If the 2-hop neighborhood is fully occupied, or if the candidate set for v_D is empty in a small or non-Eulerian component, the special-case loop can terminate without repairing degree or connectivity. Because this is the load-bearing guarantee for immediate deletion without tombstones, a rigorous proof or a redesigned algorithm with a verified invariant is required.
  2. [Algorithm 2, lines 8 and 16-24] The pseudocode for RestoreGraph is not executable as written and its control flow is unclear. Line 8 uses an undefined symbol n in 'S ← N(G,v) ∪ {n}'; this appears to be a typo for '{v}', but as written the algorithm cannot be run. In addition, the while condition on line 6 is 'c < |Vi|/2 - 1', where c counts newly reachable involved vertices, not the number of subgraphs; the accompanying text in Section 5.1 Step 4 refers to 'only d/2−1 subgraphs remain', but the Step 3 loop on lines 17-23 actually runs until |L| = 1, i.e., until all subgraphs are merged. The relationship between c, the number of subgraphs, and the stopping condition needs to be explained, and the pseudocode must be corrected, because Algorithm 2 is the specification of the paper's central correctness claim.
  3. [Section 6, Figure 2 and Table 3] All experimental results are single runs with no reported variance, error bars, or number of trials. The paper's central empirical claim—that DEG surpasses existing algorithms in construction time and search efficiency—therefore cannot be assessed for statistical reliability. For example, Table 3 reports DEG T2 at 11.0 minutes versus HNSW T2 at 23.7 minutes on SIFT1M; without multiple trials or standard deviations, it is unclear whether such margins are meaningful. Please report multiple runs or confidence intervals for the key comparisons.
minor comments (5)
  1. [Section 6.1] The phrase 'the OptimizeEdge method from crEG where used' should read 'was used'.
  2. [Figure 1 caption] The caption contains the garbled phrase 'the gray vertices form re individual a vertex induced subgraphs'; it should be 'the gray vertices form individual vertex-induced subgraphs'.
  3. [Section 5, paragraph on Even Regularity] The statement that 'a minimum regularity of 4 is required to avoid cycles and isolated vertices' is not justified as written: a 2-regular connected graph is a single cycle and has no isolated vertices. The intended property (perhaps avoiding bridges or ensuring 2-edge-connectivity) should be stated precisely.
  4. [Section 7, Figure 3] The claim that indexing and reduction time complexity 'appears logarithmic' is based on fitted curves, but no goodness-of-fit values (e.g., R²) are reported, so the reader cannot judge the quality of the fits.
  5. [Section 6.2] The description of scenario T3 is confusing: it says to build with the second half and then remove a vertex from the second half while adding a vertex from the first half; consider clarifying the intended stream direction and the resulting index contents.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: DEG's core deletion/expansion contributions are evaluated against external baselines and do not reduce to their inputs.

full rationale

The paper's central claims are the DEG vertex-deletion and graph-restoration mechanism, the distribution-agnostic expansion method, and empirical comparisons of build time, QPS, and recall. None of these claims reduces by construction to the paper's inputs or to a fitted parameter. The deletion algorithm's connectivity guarantee is asserted without a formal invariant or proof for Algorithm 2, but this is a correctness/completeness gap, not a circular derivation: the paper does not define connectivity as the output of the algorithm and then reuse that definition as evidence. The only self-citations are to the authors' prior crEG work [10] for the delta-N edge-quality metric, the OptimizeEdge parameter settings, and general design heritage; these are not used to define the measured outcomes. The evaluated results are produced against external baselines (HNSW, DiskANN, SWINN, FAISS serial scan) on standard datasets (SIFT1M, Deep1M, GloVe), and the DEG hyper-parameters are fixed constants rather than values fitted to reproduce the baselines. Consequently, no specific equation-level reduction or fitted-input-called-prediction pattern can be exhibited, and the circularity score is 0.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The central claims rely on the Eulerian even-regular graph design, on the heuristic that deleted vertices' neighbors remain close, and on six hyperparameters chosen by the authors. No new physical entities are introduced; the only new artifact is the DEG algorithm itself.

free parameters (6)
  • graph degree d = 30
    Each new vertex is connected to 30 neighbors (Section 6.1). This even-regularity parameter controls edge count and search efficiency; it is chosen by the authors and kept fixed across datasets.
  • range search result count k_ext = 60
    Used in ExtendGraph to find candidate neighbors (Section 6.1, Algorithm 3).
  • range search factor epsilon_ext = 0.1
    Used in ExtendGraph range search (Section 6.1, Algorithm 3).
  • edge optimization count k_opt = 30
    Taken from the authors' crEG paper and applied during deletion (Section 6.1).
  • edge optimization factor epsilon_opt = 0.001
    Taken from the crEG paper's OptimizeEdge settings (Section 6.1).
  • edge optimization iterations i_opt = 5
    Taken from the crEG paper, reduced by half for half-size datasets (Section 6.1).
assumptions (6)
  • standard math A connected even-regular undirected graph contains an Eulerian cycle and has no bridges, so a single edge removal cannot disconnect it.
    Invoked in Section 5 ('Connectivity') to justify the deletion design; the paper cites Toida [23] for the Eulerian construction.
  • domain assumption ANNS search can start from an entry vertex and reach any other vertex, i.e. search reachability is required.
    Section 4 lists search reachability as a required graph property for dynamic ANNS.
  • domain assumption Former neighbors of a deleted vertex tend to be close to each other in feature space and in the graph, so reconnecting them via BFS restores a useful structure.
    Section 5.1 and the ablation study rely on this proximity heuristic; no worst-case bound is given.
  • domain assumption RangeSearch returns sufficiently accurate candidate neighbors for graph extension and MRNG checking.
    Algorithm 3 uses RangeSearch to connect new vertices and to choose MRNG-conformant edges.
  • domain assumption The neighbors-of-neighbors principle holds for the constructed graphs, making BFS-based component discovery efficient.
    Section 7 states this kGraph principle as the reason BFS outperforms RangeSearch for reduction.
  • domain assumption Recall ground truth is recomputed for each dynamic scenario against the current dataset, not the original full dataset.
    Section 6 defines recall@k against the current set P, but the paper does not state how KNN ground truth was generated after deletions.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Dynamic Exploration Graph: A Novel Approach for Efficient Nearest Neighbor Search in Evolving Multimedia Datasets." pith.science (2026). https://pith.science/paper/75FQWMKB

@misc{pith2026260727640,
  author       = {Pith},
  title        = {Pith review of: Dynamic Exploration Graph: A Novel Approach for Efficient Nearest Neighbor Search in Evolving Multimedia Datasets},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/75FQWMKB}},
  note         = {Machine review of arXiv:2607.27640}
}
read the original abstract

Approximate Nearest Neighbor Search (ANNS) represents a fundamental problem in various applications (image-search, recommendation systems). While graph-based algorithms have demonstrated a good balance between search accuracy and time, handling dynamic datasets, where data points are continuously added or removed, remains a challenge. This paper introduces the Dynamic Exploration Graph (DEG), an extension of the continuous refining Exploration Graph, which retains high search efficiency for static dataset while adding essential support for dynamic data. At the core of the DEG design are two key innovations: a novel vertex deletion algorithm which guarantees graph connectivity and a data distribution-agnostic method for graph expansion. Through these mechanisms, the DEG maintains a balanced and well-connected structure, even under continuous data alterations. Empirical experiments in both streaming and online scenarios demonstrate the superior performance of the DEG, surpassing existing dynamic graph algorithms in terms of construction time and search efficiency. Although optimized for dynamic datasets, the DEG delivers results as good as current state-of-the-art approaches for static dataset, underscoring its broad applicability.

Figures

Figures reproduced from arXiv: 2607.27640 by the authors.

Figure 1
Figure 1. A 2D toy example of a DEG6. Left: The red vertex is removed, the gray vertices form there individual a vertex induced subgraphs. Middle: Reachable subgraphs are merged. Right: Sufficient paths (black solid lines) between subgraphs were identified to connect the remaining subgraphs with new edges (green solid lines). Connectivity Restoration: After vertex v and its edges are removed using Algorithm 1, each of its for… view at source ↗
Figure 2
Figure 2. The number of queries per second (QPS) in relation to recall@100 for [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗
Figure 3
Figure 3. Influence of index size on indexing speed (left) and reduction speed (mid [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

31 extracted references · 14 canonical work pages

  1. [1]

    Aguerrebere, C., Hildebrand, M., Bhati, I.S., Willke, T., Tepper, M.: Locally- adaptive quantization for streaming vector search (2024),https://arxiv.org/ abs/2402.02044

    arXiv 2024

  2. [2]

    In: ICCV

    Baranchuk, D., Douze, M., Upadhyay, Y., Yalniz, I.Z.: Dedrift: Robust similarity search under content drift. In: ICCV. pp. 10992–11001. IEEE (2023)

    work page 2023

  3. [3]

    In: Proceedings of the ACM Web Conference 2023

    Chen, P., Chang, W.C., Jiang, J.Y., Yu, H.F., Dhillon, I., Hsieh, C.J.: Finger: Fast inference for graph-based approximate nearest neighbor search. In: Proceedings of the ACM Web Conference 2023. p. 3225–3235. WWW ’23, Association for Computing Machinery, New York, NY, USA (2023). https://doi.org/10.1145/3543507.3583318

    arXiv 2023

  4. [4]

    Zeitschrift für Kristallographie - Crystalline Materials84(1-6), 109–149 (1933)

    Delaunay, B.: Neue darstellung der geometrischen kristallographie. Zeitschrift für Kristallographie - Crystalline Materials84(1-6), 109–149 (1933). https://doi.org/doi:10.1524/zkri.1933.84.1.109

    doi:10.1524/zkri.1933.84.1.109 1933

  5. [5]

    Commentarii Academiae Scientiarum Imperialis Petropolitanae8, 128–140 (1736)

    Euler, L.: Solutio problematis ad geometriam situs pertinentis. Commentarii Academiae Scientiarum Imperialis Petropolitanae8, 128–140 (1736)

    work page

  6. [6]

    Scientific Reports7(1), 12140 (Sep 2017)

    Facco, E., d’Errico, M., Rodriguez, A., Laio, A.: Estimating the intrinsic dimension of datasets by a minimal neighborhood information. Scientific Reports7(1), 12140 (Sep 2017). https://doi.org/10.1038/s41598-017-11873-y

    doi:10.1038/s41598-017-11873-y 2017

  7. [7]

    IEEE Trans- actions on Pattern Analysis and Machine Intelligence44(8), 4139–4150 (2022)

    Fu, C., Wang, C., Cai, D.: High dimensional similarity search with satellite sys- tem graph: Efficiency, scalability, and unindexed query compatibility. IEEE Trans- actions on Pattern Analysis and Machine Intelligence44(8), 4139–4150 (2022). https://doi.org/10.1109/TPAMI.2021.3067706

    arXiv 2022

  8. [8]

    Fu, C., Xiang, C., Wang, C., Cai, D.: Fast approximate nearest neighbor search with the navigating spreading-out graph. Proc. VLDB Endow.12(5), 461–474 (2019)

    work page 2019

Show all 31 references
  1. [9]

    Gao, J., Long, C.: High-dimensional approximate nearest neighbor search: with reliable and efficient distance comparison operations. Proc. ACM Manag. Data 1(2) (jun 2023). https://doi.org/10.1145/3589282

    2023 doi

  2. [10]

    In: Gurrin, C., Kongkachandra, R., Schoeffmann, K., Dang-Nguyen, D.T., Rossetto, L., Satoh, S., Zhou, L

    Hezel, N., Barthel, K.U., Schall, K., Jung, K.: An exploration graph with contin- uous refinement for efficient multimedia retrieval. In: Gurrin, C., Kongkachandra, R., Schoeffmann, K., Dang-Nguyen, D.T., Rossetto, L., Satoh, S., Zhou, L. (eds.) ICMR. pp. 657–665. ACM (2024)

    2024

  3. [11]

    IPSJ Trans

    Iwasaki, M.: Proximity search in metric spaces using approximate k nearest neigh- bor graph. IPSJ Trans. on Database3, 18–28 (2010)

    2010

  4. [12]

    In: Wallach, H., Larochelle, H., Beygelzimer, A., d'Alché-Buc, F., Fox, E., Garnett, R

    Jayaram Subramanya, S., Devvrit, F., Simhadri, H.V., Krishnawamy, R., Kadekodi, R.: Diskann: Fast accurate billion-point nearest neighbor search on a single node. In: Wallach, H., Larochelle, H., Beygelzimer, A., d'Alché-Buc, F., Fox, E., Garnett, R. (eds.) Advances in Neural ...

    2019

  5. [13]

    IEEE Trans

    Jégou, H., Douze, M., Schmid, C.: Product quantization for nearest neighbor search. IEEE Trans. Pattern Anal. Mach. Intell.33(1), 117–128 (2011)

    2011

  6. [14]

    IEEE Trans

    Li, W., Zhang, Y., Sun, Y., Wang, W., Li, M., Zhang, W., Lin, X.: Approximate nearest neighbor search on high dimensional data - experiments, analyses, and improvement. IEEE Trans. Knowl. Data Eng.32(8), 1475–1488 (2020)

    2020

  7. [15]

    IEEE Trans

    Malkov, Y.A., Yashunin, D.A.: Efficient and robust approximate nearest neighbor search using hierarchical navigable small world graphs. IEEE Trans. Pattern Anal. Mach. Intell.42(4), 824–836 (2020) 14 N. Hezel et al

    2020

  8. [16]

    Information Fusion101, 101979 (2024)

    Mastelini, S.M., Veloso, B., Halford, M., de Leon Ferreira de Car- valho, A.C.P., Gama, J.: Swinn: Efficient nearest neighbor search in sliding windows using graphs. Information Fusion101, 101979 (2024). https://doi.org/https://doi.org/10.1016/j.inffus.2023.101979

    2024

  9. [17]

    ArXiv abs/2402.04713(2024)

    Oguri, Y., Matsui, Y.: Theoretical and empirical analysis of adaptive entry point selection for graph-based approximate nearest neighbor search. ArXiv abs/2402.04713(2024)

    2024 arXiv

  10. [18]

    In: Consens, M.P., Navarro, G

    Paredes, R., Chávez, E.: Using the k-nearest neighbor graph for proximity searching in metric spaces. In: Consens, M.P., Navarro, G. (eds.) SPIRE. Lec- ture Notes in Computer Science, vol. 3772, pp. 127–138. Springer (2005). https://doi.org/doi:doi.org/10.1007/11575832_14

    2005 doi

  11. [19]

    Neural Comput

    Peng, H.: Quantization to speedup approximate nearest neighbor search. Neural Comput. Appl.36(5), 2303–2313 (aug 2023). https://doi.org/10.1007/s00521-023- 08920-3

    2023 doi

  12. [20]

    In: Proceedings of the 2014 conference on empirical methods in natural language processing (EMNLP)

    Pennington, J., Socher, R., Manning, C.D.: Glove: Global vectors for word repre- sentation. In: Proceedings of the 2014 conference on empirical methods in natural language processing (EMNLP). pp. 1532–1543 (2014)

    2014

  13. [21]

    In: Meila, M., Zhang, T

    Radford, A., Kim, J.W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., Krueger, G., Sutskever, I.: Learning transferable visual models from natural language supervision. In: Meila, M., Zhang, T. (eds.) ICML. Proceedings of Machin...

    2021

  14. [22]

    ArXiv abs/2105.09613(2021)

    Singh, A., Subramanya, S.J., Krishnaswamy, R., Simhadri, H.V.: Freshdiskann: A fast and accurate graph-based ann index for streaming similarity search. ArXiv abs/2105.09613(2021)

    2021 arXiv

  15. [23]

    Journal of Combinatorial Theory, Series B16(2), 124–133 (1974)

    Toida, S.: Construction of quartic graphs. Journal of Combinatorial Theory, Series B16(2), 124–133 (1974). https://doi.org/doi:10.1016/0095-8956(74)90054-9

    1974 doi

  16. [24]

    Pattern Recognition12(4), 261–268 (1980)

    Toussaint, G.T.: The relative neighbourhood graph of a finite planar set. Pattern Recognition12(4), 261–268 (1980). https://doi.org/doi:10.1016/0031- 3203(80)90066-7

    1980 doi

  17. [25]

    World Wide Web (WWW)26(6), 3759–3794 (November 2023)

    Ukey, N., Zhang, G., Yang, Z., Li, B., Li, W., Zhang, W.: Efficient continuous knn join over dynamic high-dimensional data. World Wide Web (WWW)26(6), 3759–3794 (November 2023)

    2023

  18. [26]

    Wang, M., Xu, X., Yue, Q., Wang, Y.: A comprehensive survey and experimen- tal comparison of graph-based approximate nearest neighbor search. Proc. VLDB Endow.14(11), 1964–1978 (jul 2021). https://doi.org/10.14778/3476249.3476255

    1964

  19. [27]

    CoRRabs/2206.10839(2022)

    Xu, Z., Zhao, W., Tan, S., Zhou, Z., Li, P.: Proximity graph mainte- nance for fast online nearest neighbor search. CoRRabs/2206.10839(2022). https://doi.org/10.48550/arXiv.2206.10839

    2022 doi arXiv

  20. [28]

    In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

    Yandex, A.B., Lempitsky, V.: Efficient indexing of billion-scale datasets of deep de- scriptors. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp. 2055–2063 (2016). https://doi.org/10.1109/CVPR.2016.226

    2016 doi

  21. [29]

    CoRR abs/2311.18724(2023)

    Yue, Q., Xu, X., Wang, Y., Tao, Y., Luo, X.: Routing-guided learned prod- uct quantization for graph-based approximate nearest neighbor search. CoRR abs/2311.18724(2023)

    2023 arXiv

  22. [30]

    Zeng, X., Wu, Z., Hu, X., Shi, X., Sun, S., Zhang, S.: Candy: A benchmark for con- tinuous approximate nearest neighbor search with dynamic data ingestion (2024), https://arxiv.org/abs/2406.19651

    2024 arXiv

  23. [31]

    Zhao, W.L., Wang, H., Ngo, C.W.: Approximate k-nn graph construction: a generic online approach (2020),https://arxiv.org/abs/1804.03032

    2020 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.