Tidy current model and datasource entries to scivision (#477)

* add thumbnail for huggingface-segmentation

* add descriptions for hugging face models

* remove "test plugin" model

* rename greenotyper model and add thumbnail

* missing commas

* model descriptions update

* rename scivision test data and add thumbnail

* update tasks

* resize Koala

* remove duplicate plankton dataset

* rename plankton dataset

* rename coastal edges

* change name

* add sentinel thumbnail and rename data

* rename plankton thumbnail

* add flowers thumbnail

* add butterfly thumbnail

* fix stardist thumbnail name

* update description coastal-edges

* add coastal to updated description coastal-edges

* update description treecrowns

* add thumbnail treecrowns

* add thumbnail coastal-edges

* add parakeet thumbnail

Co-authored-by: Alejandro © <[email protected]>

scivision/catalog/data/datasources.json

@@ -54,30 +54,12 @@
       ]
     },
     {
-      "name": "data-002",
-      "tasks": [
-        "classification"
-      ],
-      "domains": [
-        "oceanography"
-      ],
-      "url": "https://github.com/scotthosking/intake-plankton/blob/main/plktn.yaml",
-      "format": "image",
-      "labels_provided": true,
-      "institution": [
-        "CEFAS"
-      ],
-      "tags": [
-        "help-needed",
-        "2D"
-      ]
-    },
-    {
-      "name": "data-003",
-      "description": "Koala",
+      "name": "scivision-test-data",
+      "description": "Test data comprised of individual images including a Koala (suitable for classification), a busy airport (suitable for object detection), and an image containing a Dog and a Cat (suitable for segmentation).",
       "tasks": [
         "object-detection",
-        "classification"
+        "classification",
+        "segmentation"
       ],
       "domains": [
         "computer-vision"
@@ -93,7 +75,7 @@
       ]
     },
     {
-      "name": "data-004",
+      "name": "cefas-plankton",
       "description": "Sample dataset of Plankton required to demonstrate the Rapid Identification of Plankton using Machine Learning Project undertaken by CEFAS, The Alan Turing Institute and Plankton Analytics Ltd",
       "tasks": [
         "classification"
@@ -115,8 +97,8 @@
       ]
     },
     {
-      "name": "data-005",
-      "description": "Sample dataset of satellite imagery required to demonstrate the VEdge detector authored by Rogers et al. (2021)",
+      "name": "coastal-edges",
+      "description": "Sample dataset of satellite RGB imagery over two coastal locations bribie_island (RapidEye) and winterton (Planet) used to test the VEdge_Detector model described in Rogers et al (2021)",
       "tasks": [
         "classification",
         "segmentation"
@@ -141,8 +123,8 @@
       ]
     },
     {
-      "name": "data-006",
-      "description": "Sample dataset of drone RGB imagery required to demonstrate the detectreeRGB authored by Hickman et al. (2021)",
+      "name": "treecrowns",
+      "description": "Sample dataset of drone RGB imagery over Sepilok Forest Reserve, in Sabah, Malaysia, collected and processed by NERC and NEODAAS as part of Sebastian Hickman’s (2021) AI4ER MRes Project, Cambridge University",
       "tasks": [
         "object-detection",
         "segmentation"
@@ -169,7 +151,7 @@
       ]
     },
     {
-      "name": "scivision_sentinel2_stac",
+      "name": "sentinel2_stac",
       "description": "Data plugin for accessing Sentinel-2 Cloud-Optimized GeoTIFFs via scivision. The Sentinel-2 mission is a land monitoring constellation of two satellites that provide high resolution optical imagery and provide continuity for the current SPOT and Landsat missions. The mission provides a global coverage of the Earth's land surface every 5 days, making the data of great use in ongoing studies. This dataset is the same as the Sentinel-2 dataset, except the JP2K files were converted into Cloud-Optimized GeoTIFFs (COGs). Additionally, SpatioTemporal Asset Catalog metadata has were in a JSON file alongside the data, and a STAC API called Earth-search is freely available to search the archive. This dataset contains all of the scenes in the original Sentinel-2 Public Dataset and will grow as that does. L2A data are available from April 2017 over wider Europe region and globally since December 2018.",
       "tasks": [
         "classification",

scivision/catalog/data/models.json

@@ -55,7 +55,8 @@
       ]
     },
     {
-      "name": "model-002",
+      "name": "greenotyper",
+      "description": "Greenotyper is an image analysis tool for large scale plant phenotyping experiments.",
       "tasks": [
         "segmentation"
       ],
@@ -74,23 +75,6 @@
         "climate-change-and-agriculture"
       ]
     },
-    {
-      "name": "scivision-test-plugin",
-      "tasks": [
-        "object-detection"
-      ],
-      "url": "https://github.com/alan-turing-institute/scivision-test-plugin",
-      "pkg_url": "git+https://github.com/alan-turing-institute/scivision-test-plugin.git@master",
-      "format": "image",
-      "pretrained": true,
-      "labels_required": false,
-      "institution": [
-        "alan-turing-institute"
-      ],
-      "tags": [
-        "dummy"
-      ]
-    },
     {
       "name": "mapreader-plant",
       "description": "automated detection of plant patches in images of individual plants for the extraction of data on phenotypic traits",
@@ -163,6 +147,7 @@
     },
     {
       "name": "image-classifiers",
+      "description": "Scivision plugin for the image-classifiers package available at https://pypi.org/project/image-classifiers/",
       "tasks": [
         "classification"
       ],
@@ -182,6 +167,7 @@
     },
     {
       "name": "huggingface-classifiers",
+      "description": "Selection of image classifiers taken from https://huggingface.co/",
       "tasks": [
         "classification"
       ],
@@ -202,6 +188,7 @@
     },
     {
       "name": "huggingface-segmentation",
+      "description": "Selection of segmentation algorithms taken from https://huggingface.co/",
       "tasks": [
         "segmentation"
       ],
@@ -218,6 +205,7 @@
     },
     {
       "name": "huggingface-object-detection",
+      "description": "Selection of object detection algorithms taken from https://huggingface.co/",
       "tasks": [
         "object-detection"
       ],
@@ -234,6 +222,7 @@
     },
     {
       "name": "detectreeRGB-forest",
+      "description": "Detect changes in tall tree height with machine learning, LiDAR, and RGB imagery",
       "tasks": [
         "object-detection",
         "segmentation"
@@ -256,6 +245,7 @@
     },
     {
       "name": "flower-classification-model",
+      "description": "Identify flower species",
       "tasks": [
         "classification"
       ],
@@ -274,6 +264,7 @@
     },
     {
       "name": "butterfly-classification-model",
+      "description": "Identify butterfly species",
       "tasks": [
         "classification"
       ],