{ "schema_version": 1, "concept_key": "hand", "video_count": 3, "scripts": [ { "video_key": "hand_context_01", "purpose": "model", "duration_seconds": 8, "format": "16:9 landscape", "frame_rate": "25 or 30 fps constant", "resolution": "1920x1080 minimum", "reference_setup_image": "hand_context_01_setup_01.png", "reference_concept_image": "../images/hand_isolated_01.png", "story_summary": "A 5-8 second slow reveal and natural interaction with hand, keeping the target in frame.", "setting": "plain warm-white studio background", "people": "one AI-generated child body part", "target_depiction": "South Asian child's left hand shown palm forward", "timeline": [ { "timecode": "00:00.000-00:01.500", "picture": "Begin with the target partly occluded or resting naturally while its context is visible.", "performance": "Settle naturally; no one looks at camera unless the communicative meaning requires partner gaze.", "camera": "Locked establishing view; level horizon; no handheld shake.", "audio": "Capture clean room tone only.", "edit": "Start from a clean stable frame; no title card." }, { "timecode": "00:01.500-00:05.500", "picture": "Reveal or interact with the target slowly and naturally: South Asian child's left hand shown palm forward. Keep the whole target readable.", "performance": "Reveal and interact with the target once: South Asian child's left hand shown palm forward. Keep its identity, scale and appearance constant.", "camera": "Medium-close teaching view; keep the target, relevant hands, reference object and face in focus.", "audio": "Natural production sound at low level; no spoken label is required.", "edit": "One continuous take preferred; one motivated straight cut permitted." }, { "timecode": "00:05.500-00:08.000", "picture": "Hold the target fully visible for recognition and optional pause-frame use.", "performance": "Hold the completed meaning without pointing, celebration or repeated action.", "camera": "Locked result frame with at least 5 percent safe margin around the target.", "audio": "Continue matching room tone; fade neither picture nor sound before 8 seconds.", "edit": "End on a stable frame suitable for pausing." } ], "camera_direction": { "lens_equivalent": "35-50 mm natural perspective; 70-85 mm for isolated face detail", "height": "target or seated eye level", "movement": "locked-off; optional imperceptibly slow push-in", "focus": "sufficient depth of field to keep semantic evidence sharp", "exposure": "natural skin tones and no clipped highlights", "white_balance": "locked throughout", "continuity_locks": [ "participant identity", "clothing", "target instance", "background layout", "lighting direction" ] }, "performance_direction": { "primary": "Reveal and interact with the target once: South Asian child's left hand shown palm forward. Keep its identity, scale and appearance constant.", "pace": "natural and readable, never slow-motion acting", "gaze": "natural task or partner gaze; never require eye contact", "repetitions": "one complete target event", "safeguarding": "stop on discomfort; no forced compliance or distress performance" }, "spoken_script": { "required": false, "dialogue": [], "localized_voiceover_optional": "Use the approved localized semantic label or instruction for hand only after SLP and native review.", "record_separately": true }, "caption_script": { "language": "en", "picture_only_policy": "Do not caption visual action as sound. Leave captions off when the final mix contains no speech or meaningful sound.", "cues": [], "localized_caption_requirement": "Create from the locked final audio, not by translating this draft blindly." }, "audio_description": { "language": "en", "script": "In plain warm-white studio background, South Asian child's left hand shown palm forward.", "delivery": "neutral concise present tense; place in a natural pause or provide as separate accessible track", "review": "SLP and blind/low-vision accessibility review required" }, "accessibility_notes": [ "Meaning must be understandable with audio muted.", "Do not use speech, music, colour, camera movement or a transient cue as the only signal.", "Keep target visible for at least two seconds after the event.", "No flash, flicker, rapid cut, whip pan or unexpected loud sound.", "Provide player pause, replay, captions toggle and audio-description track where supported.", "Check target contrast and visibility at 320-pixel-wide playback." ], "generation_prompt": "Create one 8-second landscape 16:9 realistic clinical-educational video for the language-independent concept 'Hand' (key hand), purpose model. Use the approved reference image hand_isolated_01.png for semantic content and visual continuity, not for identity preservation. Scene: South Asian child's left hand shown palm forward in plain warm-white studio background; participants: one AI-generated child body part. Begin with a stable establishing view, show one slow natural target action or reveal, then hold the completed meaning for at least two seconds. Keep the target continuously visible, anatomically and physically plausible, safely framed and naturally lit. Use at most one simple camera cut; no fast zoom, montage or dramatic acting. Language-neutral: no necessary speech and no embedded writing. Preserve culturally respectful, disability-inclusive representation.", "negative_prompt": "no captions, labels, readable text, letters, numbers, logos, brands or watermark; no jump cuts, time lapse, fast camera movement, flicker, strobe, looping discontinuity or motion smear; no extra fingers, fused limbs, changing identity, changing clothing, morphing objects or disappearing props; no unsafe, frightening, coercive, humiliating or medically distressing behavior; no target hidden by hands, crop, foreground objects or shallow focus at the teaching moment; no music as a meaning cue and no speech required to understand the concept", "generation_settings": { "aspect_ratio": "16:9", "duration_seconds": 8, "camera_motion": "locked", "seed": "record exact provider seed when available", "reference_strength": "preserve semantics and composition; do not clone a real identity", "generate_audio": false }, "continuity_review": [ "same participant identity and anatomy", "same clothing and assistive devices", "same target object and quantity", "hands remain anatomically plausible", "background objects do not appear or disappear", "action begins and ends logically" ], "delivery": { "video_filename": "hand_context_01.mp4", "caption_filename": "hand_context_01.captions..vtt", "audio_description_filename": "hand_context_01.audio_description..wav", "poster_filename": "hand_context_01.poster.png" }, "status": "script_ready_review_required" }, { "video_key": "hand_context_02", "purpose": "functional_context", "duration_seconds": 8, "format": "16:9 landscape", "frame_rate": "25 or 30 fps constant", "resolution": "1920x1080 minimum", "reference_setup_image": "hand_context_02_setup_01.png", "reference_concept_image": "../images/hand_natural_context_01.png", "story_summary": "A 6-10 second functional everyday use of hand by a person, without unnecessary spoken language.", "setting": "simple playroom table", "people": "one AI-generated child", "target_depiction": "Black child uses the right hand to hold a red ball", "timeline": [ { "timecode": "00:00.000-00:01.500", "picture": "Begin with the target partly occluded or resting naturally while its context is visible.", "performance": "Settle naturally; no one looks at camera unless the communicative meaning requires partner gaze.", "camera": "Locked establishing view; level horizon; no handheld shake.", "audio": "Capture clean room tone only.", "edit": "Start from a clean stable frame; no title card." }, { "timecode": "00:01.500-00:05.500", "picture": "Reveal or interact with the target slowly and naturally: Black child uses the right hand to hold a red ball. Keep the whole target readable.", "performance": "Reveal and interact with the target once: Black child uses the right hand to hold a red ball. Keep its identity, scale and appearance constant.", "camera": "Medium-close teaching view; keep the target, relevant hands, reference object and face in focus.", "audio": "Natural production sound at low level; no spoken label is required.", "edit": "One continuous take preferred; one motivated straight cut permitted." }, { "timecode": "00:05.500-00:08.000", "picture": "Hold the target fully visible for recognition and optional pause-frame use.", "performance": "Hold the completed meaning without pointing, celebration or repeated action.", "camera": "Locked result frame with at least 5 percent safe margin around the target.", "audio": "Continue matching room tone; fade neither picture nor sound before 8 seconds.", "edit": "End on a stable frame suitable for pausing." } ], "camera_direction": { "lens_equivalent": "35-50 mm natural perspective; 70-85 mm for isolated face detail", "height": "target or seated eye level", "movement": "locked-off; optional imperceptibly slow push-in", "focus": "sufficient depth of field to keep semantic evidence sharp", "exposure": "natural skin tones and no clipped highlights", "white_balance": "locked throughout", "continuity_locks": [ "participant identity", "clothing", "target instance", "background layout", "lighting direction" ] }, "performance_direction": { "primary": "Reveal and interact with the target once: Black child uses the right hand to hold a red ball. Keep its identity, scale and appearance constant.", "pace": "natural and readable, never slow-motion acting", "gaze": "natural task or partner gaze; never require eye contact", "repetitions": "one complete target event", "safeguarding": "stop on discomfort; no forced compliance or distress performance" }, "spoken_script": { "required": false, "dialogue": [], "localized_voiceover_optional": "Use the approved localized semantic label or instruction for hand only after SLP and native review.", "record_separately": true }, "caption_script": { "language": "en", "picture_only_policy": "Do not caption visual action as sound. Leave captions off when the final mix contains no speech or meaningful sound.", "cues": [], "localized_caption_requirement": "Create from the locked final audio, not by translating this draft blindly." }, "audio_description": { "language": "en", "script": "In simple playroom table, Black child uses the right hand to hold a red ball.", "delivery": "neutral concise present tense; place in a natural pause or provide as separate accessible track", "review": "SLP and blind/low-vision accessibility review required" }, "accessibility_notes": [ "Meaning must be understandable with audio muted.", "Do not use speech, music, colour, camera movement or a transient cue as the only signal.", "Keep target visible for at least two seconds after the event.", "No flash, flicker, rapid cut, whip pan or unexpected loud sound.", "Provide player pause, replay, captions toggle and audio-description track where supported.", "Check target contrast and visibility at 320-pixel-wide playback." ], "generation_prompt": "Create one 8-second landscape 16:9 realistic clinical-educational video for the language-independent concept 'Hand' (key hand), purpose functional_context. Use the approved reference image hand_natural_context_01.png for semantic content and visual continuity, not for identity preservation. Scene: Black child uses the right hand to hold a red ball in simple playroom table; participants: one AI-generated child. Begin with a stable establishing view, show one slow natural target action or reveal, then hold the completed meaning for at least two seconds. Keep the target continuously visible, anatomically and physically plausible, safely framed and naturally lit. Use at most one simple camera cut; no fast zoom, montage or dramatic acting. Language-neutral: no necessary speech and no embedded writing. Preserve culturally respectful, disability-inclusive representation.", "negative_prompt": "no captions, labels, readable text, letters, numbers, logos, brands or watermark; no jump cuts, time lapse, fast camera movement, flicker, strobe, looping discontinuity or motion smear; no extra fingers, fused limbs, changing identity, changing clothing, morphing objects or disappearing props; no unsafe, frightening, coercive, humiliating or medically distressing behavior; no target hidden by hands, crop, foreground objects or shallow focus at the teaching moment; no music as a meaning cue and no speech required to understand the concept", "generation_settings": { "aspect_ratio": "16:9", "duration_seconds": 8, "camera_motion": "locked", "seed": "record exact provider seed when available", "reference_strength": "preserve semantics and composition; do not clone a real identity", "generate_audio": false }, "continuity_review": [ "same participant identity and anatomy", "same clothing and assistive devices", "same target object and quantity", "hands remain anatomically plausible", "background objects do not appear or disappear", "action begins and ends logically" ], "delivery": { "video_filename": "hand_context_02.mp4", "caption_filename": "hand_context_02.captions..vtt", "audio_description_filename": "hand_context_02.audio_description..wav", "poster_filename": "hand_context_02.poster.png" }, "status": "script_ready_review_required" }, { "video_key": "hand_context_03", "purpose": "generalisation", "duration_seconds": 8, "format": "16:9 landscape", "frame_rate": "25 or 30 fps constant", "resolution": "1920x1080 minimum", "reference_setup_image": "hand_context_03_setup_01.png", "reference_concept_image": "../images/hand_unfamiliar_01.png", "story_summary": "A 6-10 second novel setting or instance of hand for generalisation.", "setting": "simple home room with face softly behind", "people": "one AI-generated older adult", "target_depiction": "older Black woman's left hand extended palm forward", "timeline": [ { "timecode": "00:00.000-00:01.500", "picture": "Begin with the target partly occluded or resting naturally while its context is visible.", "performance": "Settle naturally; no one looks at camera unless the communicative meaning requires partner gaze.", "camera": "Locked establishing view; level horizon; no handheld shake.", "audio": "Capture clean room tone only.", "edit": "Start from a clean stable frame; no title card." }, { "timecode": "00:01.500-00:05.500", "picture": "Reveal or interact with the target slowly and naturally: older Black woman's left hand extended palm forward. Keep the whole target readable.", "performance": "Reveal and interact with the target once: older Black woman's left hand extended palm forward. Keep its identity, scale and appearance constant.", "camera": "Medium-close teaching view; keep the target, relevant hands, reference object and face in focus.", "audio": "Natural production sound at low level; no spoken label is required.", "edit": "One continuous take preferred; one motivated straight cut permitted." }, { "timecode": "00:05.500-00:08.000", "picture": "Hold the target fully visible for recognition and optional pause-frame use.", "performance": "Hold the completed meaning without pointing, celebration or repeated action.", "camera": "Locked result frame with at least 5 percent safe margin around the target.", "audio": "Continue matching room tone; fade neither picture nor sound before 8 seconds.", "edit": "End on a stable frame suitable for pausing." } ], "camera_direction": { "lens_equivalent": "35-50 mm natural perspective; 70-85 mm for isolated face detail", "height": "target or seated eye level", "movement": "locked-off; optional imperceptibly slow push-in", "focus": "sufficient depth of field to keep semantic evidence sharp", "exposure": "natural skin tones and no clipped highlights", "white_balance": "locked throughout", "continuity_locks": [ "participant identity", "clothing", "target instance", "background layout", "lighting direction" ] }, "performance_direction": { "primary": "Reveal and interact with the target once: older Black woman's left hand extended palm forward. Keep its identity, scale and appearance constant.", "pace": "natural and readable, never slow-motion acting", "gaze": "natural task or partner gaze; never require eye contact", "repetitions": "one complete target event", "safeguarding": "stop on discomfort; no forced compliance or distress performance" }, "spoken_script": { "required": false, "dialogue": [], "localized_voiceover_optional": "Use the approved localized semantic label or instruction for hand only after SLP and native review.", "record_separately": true }, "caption_script": { "language": "en", "picture_only_policy": "Do not caption visual action as sound. Leave captions off when the final mix contains no speech or meaningful sound.", "cues": [], "localized_caption_requirement": "Create from the locked final audio, not by translating this draft blindly." }, "audio_description": { "language": "en", "script": "In simple home room with face softly behind, older Black woman's left hand extended palm forward.", "delivery": "neutral concise present tense; place in a natural pause or provide as separate accessible track", "review": "SLP and blind/low-vision accessibility review required" }, "accessibility_notes": [ "Meaning must be understandable with audio muted.", "Do not use speech, music, colour, camera movement or a transient cue as the only signal.", "Keep target visible for at least two seconds after the event.", "No flash, flicker, rapid cut, whip pan or unexpected loud sound.", "Provide player pause, replay, captions toggle and audio-description track where supported.", "Check target contrast and visibility at 320-pixel-wide playback." ], "generation_prompt": "Create one 8-second landscape 16:9 realistic clinical-educational video for the language-independent concept 'Hand' (key hand), purpose generalisation. Use the approved reference image hand_unfamiliar_01.png for semantic content and visual continuity, not for identity preservation. Scene: older Black woman's left hand extended palm forward in simple home room with face softly behind; participants: one AI-generated older adult. Begin with a stable establishing view, show one slow natural target action or reveal, then hold the completed meaning for at least two seconds. Keep the target continuously visible, anatomically and physically plausible, safely framed and naturally lit. Use at most one simple camera cut; no fast zoom, montage or dramatic acting. Language-neutral: no necessary speech and no embedded writing. Preserve culturally respectful, disability-inclusive representation.", "negative_prompt": "no captions, labels, readable text, letters, numbers, logos, brands or watermark; no jump cuts, time lapse, fast camera movement, flicker, strobe, looping discontinuity or motion smear; no extra fingers, fused limbs, changing identity, changing clothing, morphing objects or disappearing props; no unsafe, frightening, coercive, humiliating or medically distressing behavior; no target hidden by hands, crop, foreground objects or shallow focus at the teaching moment; no music as a meaning cue and no speech required to understand the concept", "generation_settings": { "aspect_ratio": "16:9", "duration_seconds": 8, "camera_motion": "locked", "seed": "record exact provider seed when available", "reference_strength": "preserve semantics and composition; do not clone a real identity", "generate_audio": false }, "continuity_review": [ "same participant identity and anatomy", "same clothing and assistive devices", "same target object and quantity", "hands remain anatomically plausible", "background objects do not appear or disappear", "action begins and ends logically" ], "delivery": { "video_filename": "hand_context_03.mp4", "caption_filename": "hand_context_03.captions..vtt", "audio_description_filename": "hand_context_03.audio_description..wav", "poster_filename": "hand_context_03.poster.png" }, "status": "script_ready_review_required" } ], "approval": { "clinical": false, "cultural": false, "accessibility": false, "rights": false, "technical": false } }