FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2943407549> ?p ?o ?g. }

• W2943407549 abstract "As an instance-level recognition problem, person re-identification (ReID) relies on discriminative features, which not only capture different spatial scales but also encapsulate an arbitrary combination of multiple scales. We call features of both homogeneous and heterogeneous scales omni-scale features. In this paper, a novel deep ReID CNN is designed, termed Omni-Scale Network (OSNet), for omni-scale feature learning. This is achieved by designing a residual block composed of multiple convolutional streams, each detecting features at a certain scale. Importantly, a novel unified aggregation gate is introduced to dynamically fuse multi-scale features with input-dependent channel-wise weights. To efficiently learn spatial-channel correlations and avoid overfitting, the building block uses pointwise and depthwise convolutions. By stacking such block layer-by-layer, our OSNet is extremely lightweight and can be trained from scratch on existing ReID benchmarks. Despite its small model size, OSNet achieves state-of-the-art performance on six person ReID datasets, outperforming most large-sized models, often by a clear margin. Code and models are available at: url{https://github.com/KaiyangZhou/deep-person-reid}." @default.
• W2943407549 created "2019-05-09" @default.
• W2943407549 creator A5004087827 @default.
• W2943407549 creator A5007032481 @default.
• W2943407549 creator A5014436524 @default.
• W2943407549 creator A5030812622 @default.
• W2943407549 date "2019-05-02" @default.
• W2943407549 modified "2023-10-16" @default.
• W2943407549 title "Omni-Scale Feature Learning for Person Re-Identification" @default.
• W2943407549 cites W1799366690 @default.
• W2943407549 cites W1928419358 @default.
• W2943407549 cites W1982925187 @default.
• W2943407549 cites W2097117768 @default.
• W2943407549 cites W2108598243 @default.
• W2943407549 cites W2151065677 @default.
• W2943407549 cites W2183341477 @default.
• W2943407549 cites W2194775991 @default.
• W2943407549 cites W2204750386 @default.
• W2943407549 cites W2279098554 @default.
• W2943407549 cites W2302255633 @default.
• W2943407549 cites W2308869522 @default.
• W2943407549 cites W2410968923 @default.
• W2943407549 cites W2502225121 @default.
• W2943407549 cites W2511791013 @default.
• W2943407549 cites W2531409750 @default.
• W2943407549 cites W2549139847 @default.
• W2943407549 cites W2553498075 @default.
• W2943407549 cites W2561238782 @default.
• W2943407549 cites W2572730214 @default.
• W2943407549 cites W2585635281 @default.
• W2943407549 cites W2598634450 @default.
• W2943407549 cites W2605212061 @default.
• W2943407549 cites W2612445135 @default.
• W2943407549 cites W2620998106 @default.
• W2943407549 cites W2736410039 @default.
• W2943407549 cites W2747685395 @default.
• W2943407549 cites W2755066373 @default.
• W2943407549 cites W2769346661 @default.
• W2943407549 cites W2770739811 @default.
• W2943407549 cites W2778955544 @default.
• W2943407549 cites W2785523195 @default.
• W2943407549 cites W2798385569 @default.
• W2943407549 cites W2798458055 @default.
• W2943407549 cites W2798550112 @default.
• W2943407549 cites W2798775284 @default.
• W2943407549 cites W2798794112 @default.
• W2943407549 cites W2798874329 @default.
• W2943407549 cites W2799047769 @default.
• W2943407549 cites W2883638665 @default.
• W2943407549 cites W2884366600 @default.
• W2943407549 cites W2884585870 @default.
• W2943407549 cites W2890159224 @default.
• W2943407549 cites W2896016251 @default.
• W2943407549 cites W2896888563 @default.
• W2943407549 cites W2898047322 @default.
• W2943407549 cites W2950300355 @default.
• W2943407549 cites W2953214814 @default.
• W2943407549 cites W2954765307 @default.
• W2943407549 cites W2957037461 @default.
• W2943407549 cites W2962691289 @default.
• W2943407549 cites W2962706983 @default.
• W2943407549 cites W2962859295 @default.
• W2943407549 cites W2962926870 @default.
• W2943407549 cites W2963000559 @default.
• W2943407549 cites W2963049565 @default.
• W2943407549 cites W2963067443 @default.
• W2943407549 cites W2963078173 @default.
• W2943407549 cites W2963125010 @default.
• W2943407549 cites W2963163009 @default.
• W2943407549 cites W2963180826 @default.
• W2943407549 cites W2963263347 @default.
• W2943407549 cites W2963365374 @default.
• W2943407549 cites W2963420686 @default.
• W2943407549 cites W2963438548 @default.
• W2943407549 cites W2963446712 @default.
• W2943407549 cites W2963557071 @default.
• W2943407549 cites W2963575739 @default.
• W2943407549 cites W2963610729 @default.
• W2943407549 cites W2963690547 @default.
• W2943407549 cites W2963842104 @default.
• W2943407549 cites W2963910742 @default.
• W2943407549 cites W2963983343 @default.
• W2943407549 cites W2963993763 @default.
• W2943407549 cites W2964044605 @default.
• W2943407549 cites W2964137095 @default.
• W2943407549 cites W2964163358 @default.
• W2943407549 cites W2964304299 @default.
• W2943407549 cites W2979938149 @default.
• W2943407549 cites W2980073905 @default.
• W2943407549 cites W2981138657 @default.
• W2943407549 cites W2981684125 @default.
• W2943407549 cites W2986093954 @default.
• W2943407549 cites W3100506510 @default.
• W2943407549 cites W3118608800 @default.
• W2943407549 cites W41482161 @default.
• W2943407549 doi "https://doi.org/10.48550/arxiv.1905.00953" @default.
• W2943407549 hasPublicationYear "2019" @default.
• W2943407549 type Work @default.
• W2943407549 sameAs 2943407549 @default.
• W2943407549 citedByCount "13" @default.

• next