<tool id="sam3_semantic_segmentation" name="SAM3 Semantic Segmentation" version="1.0.0+galaxy0" profile="25.1">
    <description>
        SAM3 performs text-prompted semantic segmentation on images or videos. 
    </description>
    <requirements>
        <container type="docker">quay.io/arthur_barreau/sam3_tool:1.0.0</container>
    </requirements>
    <required_files>
        <include path="sam3_semantic_segmentation.py" />
    </required_files>
    <command detect_errors="exit_code"><![CDATA[
    mkdir data_files &&
    #set $name_file = '' 
    #if $input.input_kind == "image"
        #for $indata in $input.source
            ln -s '$indata' 'data_files/${indata.element_identifier}' &&
            #set $name_file += $indata.element_identifier + " "
        #end for
    #else
        ln -s '$input.source' 'data_files/${input.source.element_identifier}' &&
        #set $name_file = $input.source.element_identifier + " "
    #end if
    python '$__tool_directory__/sam3_semantic_segmentation.py' 
        --model '$sam3_models'
        --prompts '$text_prompt' 
        --conf '$conf' 
        --vid_stride '$vid_stride' 
        --outdir outputs 
        --outputs $outputs_format
        --name_file '$name_file'
    ]]></command>
    <inputs>
        <param name="sam3_models" label="Model data" type="select" help="Contact the administrator of our Galaxy instance if you miss model data">
            <options from_data_table="huggingface">
                <filter type="static_value" column="5" value="1"/>
            </options>
            <validator message="No model annotation is available for SAM3" type="no_options"/>
        </param>
        <conditional name="input">
            <param name="input_kind" type="select" label="Input type"
                help="Choose either multiple images or a single video (not both)">
                <option value="image">One or more images</option>
                <option value="video">One video</option>
            </param>
            <when value="image">
                <param name="source" type="data" format="jpg,png"
                    multiple="true" label="Input images"/>
            </when>
            <when value="video">
                <param name="source" type="data" format="mp4,avi,mov"
                    multiple="false" label="Input video"/>
            </when>
        </conditional>
        <param name="text_prompt" type="text" label="Text prompt" >
            <help><![CDATA[
            Comma-separated list of objects or classes to detect and segment in your images/videos.
            <br/>
            <u>Example for Wildlife:</u> <code>elephant, giraffe, lion</code>
            <br/>
            <u>Best practices:</u> Use clear, specific terms in English. Be consistent with singular/plural forms.
        ]]></help>
        </param>
        <param name="conf" type="float" value="0.25" min="0.01" max="1.0"
            label="Confidence threshold"/>
        <param name="vid_stride" type="integer" value="5" min="1" max="300"
            label="Video frame stride"
            help="For video input: process one frame every N frames."/>
        <param name="outputs_format" type="select" multiple="true" optional="true"
            label="Output formats">
            <option value="coco">COCO</option>
            <option value="yolo_bbox">YOLO bounding boxes</option>
            <option value="yolo_seg">YOLO segmentation</option>
        </param>
    </inputs>
    <outputs>
        <data name="Annotations_coco" format="json" from_work_dir="./outputs/annotations.json" label="Annotation COCO" >
            <filter>outputs_format and "coco" in outputs_format</filter>
        </data>
        <collection name="Outputs_annotated" type="list">
            <discover_datasets pattern="__name_and_ext__" directory="outputs/outputs_annotated"/>
        </collection>
        <collection name="Yolo_Bbox_Image" type="list" label="YOLO Bbox Images">
            <filter>outputs_format and 'yolo_bbox' in outputs_format</filter>
            <discover_datasets pattern="__name_and_ext__" directory="outputs/yolo_bbox/images"/>
        </collection>
        <collection name="Yolo_Bbox_Label" type="list" label="YOLO Bbox Labels">
            <filter>outputs_format and 'yolo_bbox' in outputs_format</filter>
            <discover_datasets pattern="__name_and_ext__" directory="outputs/yolo_bbox/labels"/>
        </collection>
        <collection name="Yolo_Seg_Image" type="list" label="YOLO Seg Images">
            <filter>outputs_format and 'yolo_seg' in outputs_format</filter>
            <discover_datasets pattern="__name_and_ext__" directory="outputs/yolo_seg/images"/>
        </collection>
        <collection name="Yolo_Seg_Label" type="list" label="YOLO Seg Labels">
            <filter>outputs_format and 'yolo_seg' in outputs_format</filter>
            <discover_datasets pattern="__name_and_ext__" directory="outputs/yolo_seg/labels"/>
        </collection>
    </outputs>
    <tests>
        <test expect_exit_code="1" expect_failure="true">
            <param name="sam3_models" value="unknown"/>
            <conditional name="input">
                <param name="input_kind" value="image" />
                <param name="source" value="5827603936_3f1d5d715c_z.jpg" />
            </conditional>
            <param name="text_prompt" value="elephant"/>
            <param name="conf" value="0.25"/>
            <param name="outputs_format" value="coco,yolo_bbox"/>
            <assert_stdout>
                <has_text text="Invalid model!"/>
            </assert_stdout>
        </test>
    </tests>
    <help><![CDATA[

===================================================
SAM3 – Text-Prompted Semantic Segmentation
===================================================

SAM3 allows semantic segmentation of objects in images and videos using
natural language prompts.

Typical use cases include:
- Wildlife or object segmentation in videos
- Biomedical image segmentation (cells, tissues)
- Environmental or ecological video analysis

-------------------
How it works
-------------------

1. Provide an image or video
2. Enter a text prompt describing the object to segment
3. SAM3 generates segmentation masks and overlays
4. Results are returned as images or videos

-------------------
Notes
-------------------

- For large videos, increase the frame stride to reduce processing time
    ]]></help>
    <creator>
        <person name="Arthur Barreau" email="arthurbarreau.ab@gmail.com" />
        <organization name="MOOREV" url="https://moorev.fr/" />
        <organization name="PNDB" url="https://www.pndb.fr/pages/accueil/" />
    </creator>
    <citations>
        <citation type="bibtex">
@article{sam3,
  title={SAM3: Segment Anything in Images and Videos with Language Prompts},
  author={Ultralytics},
  year={2024}
}
        </citation>
    </citations>
</tool>