ChatGPT IA Guide: Check a Garba Clapping Storyboard for Continuity

By TryBuy Editorial

Use ChatGPT to organise a garba clapping portrait storyboard into a gesture-state and garment-continuity table. Supply the three frame roles and your observations, then ask for the unresolved relationship between them. The output is a written production handoff, not a validation of choreography, anatomy or an actual performance captured in the images.

Label the three planned states

Give the frames stable IDs and describe the intended hands-approach, selected contact or near-contact, and separated-hand states. Include the authorised adult subject and kurta references. Keep the gesture description distinct from garment facts. If a sleeve or lower border is hidden by the crop, record it as unassessable rather than assuming continuity.

Supply your own candidate notes under the corresponding IDs. State whether a crop, body orientation or camera position changed intentionally. These changes affect what can be compared. The model should not interpret every visual difference as a product change, or treat a coordinated palette as proof that the same garment has been preserved.

Copy the continuity-table prompt

Convert my three-frame clapping storyboard and observation notes into a continuity table. Return frame ID, intended gesture state, observed hand relationship, intentional framing change, garment detail to compare, unresolved evidence and next action. Preserve the source labels. Do not infer a culturally correct dance step, claim physical performance testing or approve anatomy from written notes. Keep identity, gesture readability and product continuity as separate checks. Draft a narrow revision request for each reported blocker without changing already accepted details.

Ask the model to retain the difference between planned and observed. A storyboard says what the frame should communicate; it does not establish that the candidate actually does so.

Filled fictional example

Imagine Frame A and Frame C have clear written observations, while the middle frame’s hand contact is described as unclear and its cuff is partly concealed. The fictional table should request a source comparison for those areas rather than declare the gesture correct or the cuff missing. It can note that the middle frame’s role is not yet established from the supplied record.

If another note reports a changed collar in Frame C, keep that separate from the hand question. The correction should name the collar source, not rewrite the gesture sequence. This example illustrates document organisation and review priorities, not a tested model response or an actual garment error.

Revise the handoff in small steps

• “Show every conclusion that treats a planned frame role as a completed observation.”
• “Separate hidden garment detail from reported garment alteration.”
• “Draft a correction for the middle frame that preserves the original subject and approved clothing details.”

After receiving revised candidate notes, update only the affected rows. Keep the earlier version available so a new contact pose does not erase a previously logged collar concern. A traceable revision history is more useful than repeatedly renaming the whole sequence final.

Review the real candidates yourself

Check hands, wrists, visible garment construction and the actual crop against the source. View the sequence at its intended reading size as well as the relevant larger details. The text table cannot establish cultural authenticity, physical comfort or smooth real movement. It can help you decide which evidence question the next reviewer must answer.

Keep source and candidate version IDs visible in the exported handoff so another reviewer can find the exact middle frame being discussed.

Use the original practical guide for planning the three held portrait frames and the men’s kurtas collection for current product references. Your deliverable is a gesture-state table, an evidence queue and a precise revision brief. Keep concept labels attached to generated images and leave performance and product accuracy grounded in the sources and human review, rather than the fluency of the written plan.

For general instructions and follow-up techniques, see official OpenIA prompting guidance.