xahidbuffon / awesome_underwater_datasets Goto Github PK

Pointers to large-scale underwater datasets and relevant resources.

underwater-robotics marine-data underwater-images marine-resources marine-species underwater-vehicles underwater-image underwater-datasets detection stereo-data

awesome_underwater_datasets's Introduction

Image Enhancement, Color Correction/Restoration

EUVP dataset: Data, Paper, Code. (paired and unpaired data; FUnIE-GAN)
Underwater imagenet: Data, Paper, Code. (paired data; UGAN)
UIEBD dataset: Data, Paper, Code. (Water-Net)
SQUID dataset: Data, Paper, Code. (Underwater-HL)
U-45: Data, Paper. (UDAE)
RUIE benchmark: Data, Paper. (RUIE-Net)
Jamaica port royal: Data, Paper, Code. (Water-GAN)
Virtual periscope: Data, Paper.
Color correction: Data.
Color restoration: Data, Paper, Code.
TURBID data: Data, Paper.
OceanDark dataset: Data, Paper.

SISR: Single Image Super-Resolution

USR-248: Data, Paper, Code. (for 2x, 4x, and 8x training; SRDRM, SRDRM-GAN)

RGB-D: Monocular Depth Estimation

USOD10k: Data, Paper. UDepth paper

SESR: Simultaneous Enhancement and Super-Resolution

UFO-120: Data, Paper, Code. (for 2x, 3x, and 4x SESR and saliency prediction; Deep SESR)

Image Segmentation

SUIM: Data, Paper, Code. (SUIM-Net)
Coral-Net: Data, Paper, Code. (Coral-Seg)
Eilat: Data, Paper.
Change detection: Data, Paper.
LIACI: Data, Paper.

SOD: Salient Object Detection

USOD10k: Data, Paper.
UFO-120: Data, Paper, Code
MUED database: Data-1, Data-2, Paper

Object Detection/Classification

A. General

MOUSS data: Data. (CVPR 2018 workshop challenge)
MBARI databse: Data.
HabCam database: Data.
OUC-vision: Paper.
MARIS project: Data.
NOAA data: Data.
Aqualoc dataset: Data, Paper. (visual-inertial-pressure localization)
Brackish dataset: Data, Paper.
SUN database (underwater scenes): Data.
FathomNet (image database): Data.

B. Human-robot cooperation

Diver detection: Data, Paper.
Robot tracking by detection: Data, Paper.
CADDY diver pose data: Data, Paper.

C. Coral-reef

Moorea corals (UCSD): Data, Paper.
Coral-reef Puerto Rico: Data.
Coral-Net: Data.

D. Fish

WildFish database: Data, Paper.
Labeled fishes: Data, Paper.
Fish4Knowledge data: Data.
Fish database: Data.
AQUALIFEIMAGES database: Data.
Rockfish: Data.
Fish recognition data: Data, Paper.
Oceanwide images: Data.
Fish detection and tracking: Data, Paper.
Fish trajectory detection: Data, Paper.

E. Trash and marine debris

TrashCan: Data, Paper
Trash-ICRA19: Data, Paper
Deep-sea debris database: Data, Paper.
Tiny plastics posing threat to turtles: Data, Paper.

6. Acoustic Data

Five-element acoustic dataset: Data, Paper.
DIDSON dataset: Data1, Data2, Data3, Paper. (fishery classification and assessment)
Spectrogram Analysis: Data, Paper.
Caves sonar and vision data: Data, Paper.

Stereo Data

Tasmania coral point, Scott reef-25, O'Hara-7: Data, Paper.
Stereo from Flicker: Data, Paper.
CADDY stereo data: Data, Paper.
HIMB data for UWStereoNet: Data, Paper. (UW-StereoNet)
SQUID dataset: Data, Paper

Docking Data

Underwater Docking Images Dataset(UDID): Data, Paper.

Temperature Data

Underwater temperature dataset: Data.

awesome_underwater_datasets's People

Stargazers

Watchers

Forkers

peigenluo feng135799 apaeslin mehtaz23 mikkiewwkw bilityniu ipnuistlegal ffjjff813019 zhenqifu librechou dbofseuofhust darkmatter18 kostas-alexis damianbrzoza veechiry langtaoli lele-zhou alzayats giuliadm79 hanruye sun-jinghao robot-learning-group xxzcc owen718 wushengcong yangapple-netizen gopirajumatta sia-underwater-robots eunpil-park shumailaahmed liqianqian123 rainfalj yudadabing research-mrutyunjaya perseverancelx mmehdimousavi rch-perfect ashishd huyq19 mksarker frensher liuxi1997 zhouda98 rikrd ouyaming julingers cvlzw 1349115419 sih-22-kyogre alexabngn zhipengy liamlian0727 i4737 abelzw rockwenjj wangdeqing thithaotran electronicsarchiver liuqian564684 duweihua deasonyuan way1122 zhuangpeixian hippymitty aws-ahmad xuridongshengpxd xxccb littlecarrotz riponazad jieblue mariamalz cn-github007 googolplexgoodenough linzha0 steven-yang-yds bbbbbh-cmd joeypjx awhitewhale cvmll-awesome zhanghansenit ahmedharbii bjersgen iam-007swarna kyriema 1716932646 hitzhangyu alloblue0 xiaohaierguang simonlove0513 bmaser wjhsmn424896 xuanjiawang huanyu2019 mobi1019 yokko123 1085437068 h-wenfeng bigwangyudong jiamincheng wrqf

awesome_underwater_datasets's Issues

I would like to ask if there is a publicly available RGB-D underwater datasets

Since underwater depth information is needed in my experiment, I would like to ask whether there is an RGB-D underwater data set

Some issues about the SUIM

https://github.com/xahidbuffon/SUIM
I downloaded the SUIM dataset on your lab websites. I trained on traditional u-net (not the code you released). I know that about 83 images are different from the size of the mask. And I deleted these images. I process masks to make sure they are labeled correctly and can be entered into the network. eg: [255,255,255] - >[1,1,1]. To make sure these inputs are correct, I also check the pictures by visualizing these datasets. After that, I train it and set the input size(320,240) epoch 30, steps 5000 batch size 2. But I could not get the result as you mentioned in the paper. I got 10-20% fewer results than the different types of IOU in the paper.

After that, I trained on the code you released. And I found that the data loader is slow, in this case, I used logical operations to speed up the process. Here is my code.

def getSaliency(mask):
    # one combined category: HD/RO/FV/WR
        
    """
    RGB color code and object categories:
    ------------------------------------
    000 BW: Background waterbody
    001 HD: Human divers
    010 PF: Plants/sea-grass
    011 WR: Wrecks/ruins
    100 RO: Robots/instruments
    101 RI: Reefs and invertebrates
    110 FV: Fish and vertebrates
    111 SR: Sand/sea-floor (& rocks)
    """
    imw, imh = mask.shape[0], mask.shape[1]
    sal = np.zeros((imw, imh))
    # bkg = np.zeros((imw,imh))
    a = np.logical_and(np.logical_and(np.logical_not(mask[:, :, 0]),np.logical_not(mask[:,:, 1])),mask[:, :, 2])
    sal[a] = 1
    a = np.logical_and(np.logical_and(np.logical_not(mask[:, :, 2]), np.logical_not(mask[:, :, 1])), mask[:, :, 0])
    sal[a] = 1
    a = np.logical_and(np.logical_and(mask[:,:,0],mask[:,:,1]),np.logical_not(mask[:,:,2]))
    sal[a] = 1
    a = np.logical_and(np.logical_and(mask[:, :, 1], mask[:, :, 2]), np.logical_not(mask[:, :, 0]))
    sal[a] = 1

    return np.expand_dims(sal, axis=-1)


def getRobotFishHumanReefWrecks(mask):
    # for categories: HD, RO, FV, WR, RI
    imw, imh = mask.shape[0], mask.shape[1]
    mask = np.rint(mask)
    Human = np.zeros((imw, imh))
    Robot = np.zeros((imw, imh))
    Fish = np.zeros((imw, imh))
    Reef = np.zeros((imw, imh))
    Wreck = np.zeros((imw, imh))
    a = np.logical_and(np.logical_and(np.logical_not(mask[:, :, 0]),np.logical_not(mask[:,:, 1])),mask[:, :, 2])
    Human[a] = 1
    a = np.logical_and(np.logical_and(np.logical_not(mask[:, :, 2]), np.logical_not(mask[:, :, 1])), mask[:, :, 0])
    Robot[a] = 1
    a = np.logical_and(np.logical_and(mask[:,:,0],mask[:,:,1]),np.logical_not(mask[:,:,2]))
    Fish[a] = 1
    a = np.logical_and(np.logical_and(mask[:, :, 0], mask[:, :, 2]), np.logical_not(mask[:, :, 1]))
    Reef[a] = 1
    a = np.logical_and(np.logical_and(mask[:, :, 1], mask[:, :, 2]), np.logical_not(mask[:, :, 0]))
    Wreck[a] = 1
    return np.stack((Robot, Fish, Human, Reef, Wreck), -1) �

However, I did not get the same size model as you released on google drive. After I check on the Internet, I change
save_weights_only = True
to
save_weights_only = False
But my model is still 2KB smaller than yours. And the result I tested was still bad. I got 10-20% fewer results than the different types of IOU in the paper. To improve the result, I change the batch size from 2 to 32, but similarly, the result is still bad.