Egocentric Gift Wrapping Video Data

Egocentric gift wrapping data is real-world video of a person wrapping an item in paper and folding it. The footage shows the hands, the item and the sheet as the folds are made. Alongside the video, each episode carries depth, 3D hand pose, contact events and frame-accurate action segments, all aligned on one hardware clock.

Teams search for it under several names, including gift wrapping robot training data, a paper wrapping video dataset and egocentric wrapping and taping data. They all describe demonstrations of a deformable task, collected to a written spec. This page covers only what the Firsthand site states about it and marks everything else as Not stated.

The site states one wrapping task in its commercial task catalog. The row is named "Wrap + fold packaging" and described as "Wrap an item in paper and fold". That wording names two operations: wrapping an item in paper and folding it.

Anything outside that wording is Not stated. That includes taping the paper shut, tying or curling ribbon, adding bows or tags and wrapping in tissue, so none is claimed here. If your policy needs one of them, you write it into the spec and we scope it, because the catalog row is a starting vocabulary, not a fixed menu.

Not by name. The catalog words the row as packaging, not as a gift, and the packing and kitting page files it under "Bagging and wrapping" next to a bag-and-tag row. The word gift in the route name is the search term teams use. What the site states is the paper-wrapping task itself: wrap an item in paper and fold.

A gift is a purpose, not a different hand sequence in the catalog, so a team that wants gift-specific steps writes them into the spec. Ribbon, bows, tags and decorative finishes are Not stated, and no readable gift tag or message is described on this site.

The wrap and fold row is rated Medium. It sits in the Manufacturing & Workshops sector, in an environment called Deformables & Textiles Bench, and it lists a multiview rig with handheld grippers as the capture mode. Its status is Available on request.

The site defines that status plainly: the task is scoped and recruited to spec when a buyer asks for it. That means recording for the row is not already running, so a first batch is quoted against your spec. The site does not state how many hours, episodes or contributors exist for it.

The Deformables & Textiles Bench lists a cloth-and-cable bench with garments, towels, cables, form boards and bags among its examples. It holds 12 rows. Three are rated Easy, six Medium and three Hard, and the wrap and fold row is one of the Medium six. All 12 are Available on request, and all 12 list the multiview rig with handheld grippers.

The site defines that capture mode as a staged working set: a multiview rig records the scene while the operator performs the task with handheld grippers, so demonstrations map onto a robot end effector. So for this row the site describes a staged set, not a real workplace. It does not list the head-mounted phone mode for the row, which the site describes as real work on site.

The site names no shop, store, retailer or venue, and does not state the number of sites, the regions, the room or store types or the kinds of item. It does not state capture at a customer site either. Environment, access and schedule are agreed in the written spec.

A multiview rig records the staged working set while the operator performs the task with handheld grippers, so the demonstration maps onto a robot end effector. Camera count, resolution, frame rate and bench layout are Not stated. Because the row lists the rig only, a head-mounted first-person view of this row is not stated either, so raise it in the spec if your policy perceives from a head or wrist camera.

The site does state the viewpoint argument in general: a robot or humanoid perceives from a head- or wrist-mounted camera, not a static overhead shot, and egocentric capture matches that geometry. Which streams are populated in the gripper mode is confirmed in the spec.

Each episode follows the Firsthand episode schema: ten streams synced under 2 ms, with RGB, depth, 3D hand pose with per-joint visibility, contact and grasp events, and frame-accurate action segments. For wrapping, hand pose tracks the fingers and wrist as they hold and crease the sheet, contact events mark when a hand or gripper meets or leaves the item or paper, and action segments mark each step of the job.

Object labels use the terms your spec sets. Paper types, ribbon types, item sizes and counts, package counts, tools and room types are Not stated. Task durations, cycle times, throughput and the number of episodes are Not stated too, as are success or quality labels, certifications and robot results.

The site rates the row Medium, one of six Medium rows among the 12 in its environment. That rating is the only statement on difficulty. The site gives no technical analysis of paper wrapping, so this page makes no claim about robot results, and it does not state how depth behaves on thin or reflective sheets.

What the row wording and the environment imply is a deformable object. The Deformables & Textiles Bench groups wrapping with folding, flattening, smoothing, bagging and similar sheet and cloth tasks, and the site has a separate page on deformable object data. A sheet changes shape with every touch, so the state of the task is carried by the hands, the item and the sheet together, which is why a spec names the start and end state of each fold.

A good spec treats each named operation as its own task. Wrapping the item in paper and folding it each get a start state, an end state and a set of accept criteria. That keeps action segments consistent across contributors and lets you ask for more of one step without re-collecting the others.

Finishing steps such as taping, ribbon or tags can be added as further tasks. The site does not state them, so they are scoped and quoted rather than assumed.

Other pages cover other task families: packing and kitting, box and parcel handling, deformable object manipulation, knot tying and rope handling and laundry folding. Each describes a different row or group of rows in the same task catalog.

The sibling rows on the same bench are separate tasks, for example bagging and sealing soft goods and folding shirts and towels. This page centres on the one wrap and fold row. Pallet wrapping with stretch film is a different task with its own page, not paper wrapping.

Delivery formats are RLDS, LeRobot and WebDataset, plus HDF5, zarr and Rerun .rrd. Wrapping episodes use the same schema as every other Firsthand collection, so a loader that reads another Firsthand dataset reads this one without changes. The format table below sets each format against what its official documentation states.

Every batch is validated against signed accept criteria before it ships, with a reject log, and delivery is rolling, per validated batch. Rejected items are recaptured at our cost, and billing is per validated hour.

Not in standard capture. Force or tactile readings are Not stated for the wrap and fold row. The site separately offers a tactile glove tier as a pilot-ready option, in which glove signals are time-synced to a head-mounted stereo camera by hardware timestamps and an IMU. The site does not state that the tier has been used on wrapping, so agree the fit with us first.

Every episode links to a signed contributor release through a documented consent chain, and the consent record is part of the delivery. A vision model runs on every frame at ingest, flagging faces, heads, licence plates and any legible screen or document. Those regions are irreversibly blurred, and a human reviewer confirms the result before footage ships.

If the reviewer is unsure a region is fully covered, the episode is held or discarded, never downgraded and shipped. The site does not state that readable signage, labels or gift tags are blurred, so raise any such need before the spec is signed. It states no in-home capture for this work.

Where coverage is already deep, the first validated batch lands in a median of 48 hours from signed spec. Thin environments typically take three to four weeks, because vetting operators there is the bottleneck, and we quote that up front. The wrap and fold row is Available on request, so it is recruited to your spec first, and the 48 hour figure should not be assumed.

The site also states program milestones. A pilot reaches its first delivery in two to three weeks, and a standing program reaches full rate in four to six weeks. Those assume capture coverage is already in place. On-demand top-ups run against the same spec and accept criteria after each training run. A batch size and a later-batch turnaround are Not stated, so both are quoted against your spec.

You describe the wrapping operations, the items and the sequence your policy needs. We turn that into a written spec covering tasks, environments, capture tier, annotation depth and accept criteria, then quote reach, timeline and price before any capture begins. This is the custom egocentric data collection route that every Firsthand task family follows.

A Spec Pilot, a small and fast first batch, can prove the spec and accept criteria before you scale. Every collection ships under a perpetual, buyer-owned commercial license, with a reject log and signed accept criteria on every batch.