Upload croissant.json

croissant.json

@@ -0,0 +1,113 @@
+{
+  "@context": {
+    "@language": "en",
+    "@vocab": "https://schema.org/",
+    "arrayShape": "cr:arrayShape",
+    "citeAs": "cr:citeAs",
+    "column": "cr:column",
+    "conformsTo": "dct:conformsTo",
+    "containedIn": "cr:containedIn",
+    "cr": "http://mlcommons.org/croissant/",
+    "data": {
+      "@id": "cr:data",
+      "@type": "@json"
+    },
+    "dataBiases": "cr:dataBiases",
+    "dataCollection": "cr:dataCollection",
+    "dataType": {
+      "@id": "cr:dataType",
+      "@type": "@vocab"
+    },
+    "dct": "http://purl.org/dc/terms/",
+    "extract": "cr:extract",
+    "field": "cr:field",
+    "fileProperty": "cr:fileProperty",
+    "fileObject": "cr:fileObject",
+    "fileSet": "cr:fileSet",
+    "format": "cr:format",
+    "includes": "cr:includes",
+    "isArray": "cr:isArray",
+    "isLiveDataset": "cr:isLiveDataset",
+    "jsonPath": "cr:jsonPath",
+    "key": "cr:key",
+    "md5": "cr:md5",
+    "parentField": "cr:parentField",
+    "path": "cr:path",
+    "personalSensitiveInformation": "cr:personalSensitiveInformation",
+    "recordSet": "cr:recordSet",
+    "references": "cr:references",
+    "regex": "cr:regex",
+    "repeated": "cr:repeated",
+    "replace": "cr:replace",
+    "sc": "https://schema.org/",
+    "separator": "cr:separator",
+    "source": "cr:source",
+    "subField": "cr:subField",
+    "transform": "cr:transform",
+    "rai": "http://mlcommons.org/croissant/RAI/",
+    "prov": "http://www.w3.org/ns/prov#"
+  },
+  "@type": "sc:Dataset",
+  "distribution": [
+    {
+      "@type": "cr:FileObject",
+      "@id": "repo",
+      "name": "repo",
+      "description": "The Hugging Face git repository.",
+      "contentUrl": "https://huggingface.co/datasets/madivanhorn/cfc26/tree/refs%2Fconvert%2Fparquet",
+      "encodingFormat": "git+https",
+      "sha256": "https://github.com/mlcommons/croissant/issues/80"
+    }
+  ],
+  "recordSet": [],
+  "conformsTo": "http://mlcommons.org/croissant/1.1",
+  "name": "cfc26",
+  "description": "madivanhorn/cfc26 dataset hosted on Hugging Face and contributed by the HF Datasets community",
+  "alternateName": [
+    "madivanhorn/cfc26"
+  ],
+  "creator": {
+    "@type": "Person",
+    "name": "Madison Van Horn Mitchell",
+    "url": "https://huggingface.co/madivanhorn"
+  },
+  "keywords": [
+    "mit",
+    "🇺🇸 Region: US"
+  ],
+  "license": "https://choosealicense.com/licenses/mit/",
+  "url": "https://huggingface.co/datasets/madivanhorn/cfc26",
+  "rai:dataLimitations": "The dataset is limited to sonar-based fish counting in riverine environments and may not generalize to other sensing modalities (e.g., optical video), marine/open-ocean settings, or species with substantially different motion or morphology. While expanded to nine river systems, geographic and ecological coverage remains incomplete (North America only), and certain river conditions (e.g., extreme turbidity, rare seasonal events, or atypical flow regimes) may be underrepresented. Annotation quality may be affected by low signal-to-noise ratios and ambiguity in fish visibility, particularly in dense or occluded scenes. The dataset is not recommended for tasks requiring fine-grained species classification, individual fish re-identification, or appearance-based recognition independent of motion cues.",
+  "rai:dataBiases": "The dataset may exhibit selection bias toward rivers and deployment sites where sonar systems and data collection is feasible, potentially over-representing current sites of interest in North America. Environmental conditions (e.g., typical flow, sonar artifacts) and species distributions may be unevenly sampled, leading to skewed representations of fish behavior and density. Label bias may arise from human annotation under low-visibility conditions, where temporal cues are required and ambiguous cases may be inconsistently labeled. These biases may cause models to overfit to common environmental patterns or fish movement dynamics and underperform in rare or extreme conditions.",
+  "rai:personalSensitiveInformation": "The dataset does not contain personal or human-related sensitive information. It consists of sonar imagery of fish in natural river environments.",
+  "rai:dataUseCases": "The dataset represents fish passage and abundance as measured through sonar-based observations, with the goal of estimating population counts in river systems. It is intended for benchmarking detection and counting models under realistic deployment conditions, including generalization to unseen locations. Validated use cases include: (1) evaluation of fish counting algorithms under in-distribution and out-of-distribution settings, (2) development of models for ecological monitoring and fisheries management, and (3) analysis of robustness to environmental variation and dataset imbalance. These use cases are supported by prior work on CFC22 and its deployment in real-world counting systems. The dataset is not validated for unrelated vision tasks (e.g., generic object detection), nor for ecological inference beyond aggregate counting (e.g., behavioral studies or species classification).",
+  "rai:dataSocialImpact": "Positive impacts include improved accuracy and reliability of fish population estimates, which can support sustainable fisheries management, conservation efforts, and evidence-based environmental policy. By enabling evaluation under distribution shift, the dataset may reduce deployment failures in new locations. Potential risks include over-reliance on automated systems without sufficient human oversight, which could propagate systematic errors into policy decisions. There is also a risk that models trained on this dataset may perform poorly in underrepresented environments, leading to inequities in monitoring quality across regions. No direct risks to individuals or communities are present, and no access restrictions are currently imposed beyond standard research use.",
+  "rai:hasSyntheticData": false,
+  "prov:wasDerivedFrom": [
+    {
+      "@id": "https://data.caltech.edu/records/1y23m-j8r69",
+      "prov:label": "The Caltech Fish Counting Dataset (CFC22)",
+      "sc:license": "MIT"
+    }
+  ],
+  "prov:wasGeneratedBy": [
+    {
+      "@type": "prov:Activity",
+      "prov:type": "Data Collection",
+      "prov:label": "Collection",
+      "prov:description": "Data was collected via sonar camera deployments in river systems across four U.S. states and two countries. The collection method is observations using fixed-position sonar sensors capturing continuous video streams over extended durations under varying environmental conditions."
+    },
+    {
+      "@type": "prov:Activity",
+      "prov:type": "Preprocessing",
+      "prov:label": "Preprocessing",
+      "prov:description": "Video streams were segmented into clips, filtered for quality, and normalized for consistent formatting. Clips with unusable signal quality or severe artifacts may have been excluded. Standardization steps include temporal clipping, echogram visualizations, and formatting for downstream annotation and evaluation."
+    },
+    {
+      "@type": "prov:Activity",
+      "prov:type": "Annotation",
+      "prov:label": "Annotation",
+      "prov:description": "Human annotators labeled fish instances using bounding boxes, with counting derived from crossing events. Annotation guidelines emphasized temporal consistency and handling of ambiguous detections in low signal-to-noise conditions. Quality control included review processes and consistency checks across annotators; inter-annotator agreement metrics are not reported."
+    }
+  ]
+}