Dataset Card for WHOI-Plankton - A Large Scale Fine Grained Visual Recognition Benchmark Dataset for Plankton Classification

The data set available here comprises > 3.5 million images of microscopic marine plankton, organized according to category labels provided by researchers at the Woods Hole Oceanographic Institution (WHOI). The images are currently placed into one of 103 categories.

• Original dataset available online at: https://github.com/hsosik/WHOI-Plankton.
• Original dataset license: .

Details

• train split means (RGB): [0.5041058480453973, 0.5041527920081967, 0.5042772698612863]
• train split standard deviations (RGB): [0.22517094850071867, 0.2252425962532338, 0.22524934255114173]

Samples per class for split train

0: Akashiwo                         5.00
1: Amphidinium_sp                   210.00
2: Asterionellopsis                 1853.00
3: Bacillaria                       13.00
4: Bidulphia                        31.00
5: Cerataulina                      21686.00
6: Cerataulina_flagellate           142.00
7: Ceratium                         784.00
8: Chaetoceros                     ▇ 46957.00
9: Chaetoceros_didymus              502.00
10: Chaetoceros_didymus_flagellate  1427.00
11: Chaetoceros_flagellate          149.00
12: Chaetoceros_other               276.00
13: Chaetoceros_pennate             970.00
14: Chrysochromulina                515.00
15: Ciliate_mix                     15197.00
16: Cochlodinium                    14.00
17: Corethron                       6386.00
18: Coscinodiscus                   711.00
19: Cylindrotheca                   20832.00
20: DactFragCerataul                5906.00
21: Dactyliosolen                   13020.00
22: Delphineis                      401.00
23: Dictyocha                       2358.00
24: Didinium_sp                     24.00
25: Dinobryon                       7543.00
26: Dinophysis                      357.00
27: Ditylum                         5161.00
28: Ditylum_parasite                258.00
29: Emiliania_huxleyi               148.00
30: Ephemera                        633.00
31: Eucampia                        2191.00
32: Euglena                         764.00
33: Euplotes_sp                     25.00
34: G_delicatula_detritus           816.00
35: G_delicatula_external_parasite  433.00
36: G_delicatula_parasite           2965.00
37: Gonyaulax                       564.00
38: Guinardia_delicatula           ▇ 38268.00
39: Guinardia_flaccida              1290.00
40: Guinardia_striata               3234.00
41: Gyrodinium                      677.00
42: Hemiaulus                       18.00
43: Heterocapsa_triquetra           1413.00
44: Karenia                         4.00
45: Katodinium_or_Torodinium        395.00
46: Laboea_strobila                 1470.00
47: Lauderia                        295.00
48: Leegaardiella_ovalis            245.00
49: Leptocylindrus                 ▇▇ 125690.00
50: Leptocylindrus_mediterraneus    392.00
51: Licmophora                      313.00
52: Mesodinium_sp                   3464.00
53: Odontella                       86.00
54: Paralia                         666.00
55: Parvicorbicula_socialis         78.00
56: Phaeocystis                     2388.00
57: Pleuronema_sp                   115.00
58: Pleurosigma                     2267.00
59: Prorocentrum                    4011.00
60: Proterythropsis_sp              665.00
61: Protoperidinium                 78.00
62: Pseudochattonella_farcimen      283.00
63: Pseudonitzschia                 4307.00
64: Pyramimonas_longicauda          616.00
65: Rhizosolenia                   ▇ 40632.00
66: Skeletonema                     17454.00
67: Stephanopyxis                   57.00
68: Strobilidium_morphotype1        250.00
69: Strombidium_capitatum           47.00
70: Strombidium_conicum             87.00
71: Strombidium_inclinatum          222.00
72: Strombidium_morphotype1         726.00
73: Strombidium_morphotype2         311.00
74: Strombidium_oculatum            152.00
75: Strombidium_wulffi              157.00
76: Thalassionema                   5639.00
77: Thalassiosira                   13139.00
78: Thalassiosira_dirty             2183.00
79: Tiarina_fusus                   13.00
80: Tintinnid                       2874.00
81: Tontonia_appendiculariformis    46.00
82: Tontonia_gracillima             361.00
83: amoeba                          219.00
84: bad                             9964.00
85: bead                            395.00
86: bubble                          19.00
87: clusterflagellate               290.00
88: detritus                       ▇▇▇▇▇▇▇ 378501.00
89: diatom_flagellate               728.00
90: dino30                         ▇ 47718.00
91: dino_large1                     180.00
92: flagellate_sp3                  1905.00
93: kiteflagellates                 525.00
94: mix                            ▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 2606720.00
95: mix_elongated                  ▇ 67325.00
96: other_interaction               2937.00
97: pennate                         5313.00
98: pennate_morphotype1             238.00
99: pennates_on_diatoms             851.00
100: pollen                         19.00
101: spore                          381.00
102: zooplankton                    63.00

Reference

Orenstein, E. C., Beijbom, O., Peacock, E. E., & Sosik, H. M. (2015). WHOI-Plankton - A Large Scale Fine Grained Visual Recognition Benchmark Dataset for Plankton Classification. CoRR, abs/1510.00745. http://arxiv.org/abs/1510.00745

BibTEX

@article{dataset:whoi,
  author       = {Eric C. Orenstein and Oscar Beijbom and Emily E. Peacock and Heidi M. Sosik},
  title        = {WHOI-Plankton - {A} Large Scale Fine Grained Visual Recognition Benchmark
                  Dataset for Plankton Classification},
  journal      = {CoRR},
  volume       = {abs/1510.00745},
  year         = {2015},
  url          = {http://arxiv.org/abs/1510.00745},
  eprinttype   = {arXiv},
  eprint       = {1510.00745},
}

Usage

from datasets import load_dataset

dataset = load_dataset("project-oceania/whoi-plankton")