SWISSAI DATA as of 2026-07-22 14:14

#5

/capstor/store/cscs/swissai/infra01/vision-datasets/raw/stage2/hf___ai4ce___EgoPAT3Dv2___cleaned
kindwebdataset
statusactive
samplesn/a
size3.6 GB
files510
first seen2026-07-22 13:20
last seen2026-07-22 13:20
registered2026-07-22 13:20

samples

#keyimg1.jpgimg2.jpgimg3.jpgtxt
1
1_1_1_000039
Action Summary: Right hand grasps the "C++" book (technical book, white cover with red/black text) on the top shelf via a power grasp, then lifts it toward the left.  
Initial State: The "C++" book rests on the top shelf’s right section; its cover features red/black text and a red tag. A hand approaches from the bottom right, fingers extended to grasp the book’s top cover.  
Kinematics & Contact: In Frame 1, the hand nears the book. In Frame 2, a power grasp (fingers wrapping the top cover) establishes contact. In Frame 3, the book is lifted, with the hand’s grip maintaining contact as it moves upward and leftward.  
Final State: The "C++" book is raised off the shelf, held in a maintained power grasp; placement is pending beyond Frame 3.  
Context: The action occurs on a black bookshelf with other books below/around; no tools or environmental changes beyond the book’s repositioning.
2
1_1_1_000077
Action Summary: Right hand performs a pick action, grasping the C++ book from the shelf and moving it toward the upper shelf area.  
Initial State: At Frame 1, the C++ book (category: book, size: medium, shape: rectangular with white cover and red bottom section, material: paper/paperback) is positioned upright in the right section of a black bookshelf. The right hand is positioned to grasp the book’s lower portion (base), with fingers wrapped around the red section, thumb supporting from the side.  
Kinematics & Contact: In Frame 1, the hand applies a power grasp (likely palmar/hook) with the right hand, making contact at the book’s base (red section). The hand lifts the book upward and slightly backward (away from the shelf) through Frame 2, maintaining a secure grip. Contact remains constant, with no release or adjustment in grip. In Frame 3, the hand continues to hold the book, now elevated and positioned above the other books on the shelf, still grasping the base.  
Final State: The C++ book is now elevated above the other books on the shelf, held securely by the right hand, with the grasp maintained and contact at the base.  
Context: The action occurs on a black bookshelf; other objects include a Monet book (stacked below the C++ book) and various items in lower compartments. The hand moves the C++ book from its original shelf position to a higher shelf position.
3
1_1_1_000115
Action Summary: Right hand grasps the "C++ Primer+" book and lifts it away from the black shelf.  
Initial State: At Frame 1, a right hand is approaching a black shelf. On the shelf, there is a white book with a red and gray "C++ Primer+" cover, positioned next to a blue - spined book with a white cover. The hand is positioned laterally to the left of the “C++ Primer+” book, oriented to grasp the book’s top edge.  
Kinematics & Contact: From Frame 1 to Frame 2, the hand establishes a power grasp around the lateral face of the “C++ Primer+” book, with contact primarily at the top edge of the book cover. The hand then lifts the book upward and slightly forward, moving it away from the shelf. As the book is lifted, the contact point shifts to the book’s top edge while being carried. The grip remains consistent, supporting the book’s weight.  
Final State: At Frame 3, the “C++ Primer+” book is no longer in contact with the shelf; it is elevated in the hand, oriented horizontally as the hand moves it away from the shelf.
4
1_1_1_000153
Action Summary: Right hand picks up the book from the shelf and lifts it away from the shelf.  
Initial State: In Frame 1, a right hand is distal, reaching toward the “C++ Primer” book on the top shelf. Other books (including a blue book and a white book with a red pattern) are positioned on the left side of the shelf. The top shelf has a dark surface, and the hand approaches the book’s spine (base).  
Kinematics & Contact: In Frame 1, the hand makes contact with the edge of the book (base). In Frame 2, the hand lifts the book, moving it upward and slightly forward while maintaining contact with the spine. In Frame 3, the book is fully lifted, with the hand gripping the spine, and the hand moves the book away from the shelf.  
Final State: The book is now in the right hand, lifted away from the top shelf. Grasp is maintained, and the book is oriented vertically.  
Context: The action involves a human hand interacting with books on a dark - colored shelf. No other tools or significant environment changes; the focus is on the book - to - hand movement.
5
1_1_1_000191
Action Summary: Right hand performs a pick and place action, grasping a black rectangular object and placing it onto the shelf.  

Initial State: Frame 1 shows a right hand (proximal to the camera, distal near the shelf) reaching toward a stack of books. The stack includes a white book with a red patterned top section, a blue book with text, and a lower book with "C++ Primer" visible; the hand approaches the topmost black rectangular object on this stack.  

Kinematics & Contact: In Frame 1, the hand makes contact with the top surface of the black rectangular object using a power grasp (fingers and thumb wrapping around the object’s top). The hand lifts the object, moving it upward and slightly forward. In Frame 2, the hand repositions the object above the stack, and in Frame 3, the hand releases the object, which now rests on the stack.  

Final State: The black rectangular object is stationary on the shelf, placed atop the stack of books that includes “C++ Primer.” The hand has released the object, and grasp is no longer maintained.  

Context: No tools are used. The action involves a single-hand, bimanual placement (hand approaching, lifting, and placing the object). No other objects enter/leave the action’s focus, and the environment (bookshelf) remains unchanged.
6
1_1_1_000234
Action Summary: Right hand picks up the book from the top shelf, transitioning from the left side of the shelf to an intermediate position, and lifting the book away from the shelf. 

Initial State: At Frame 1, a black bookshelf holds multiple objects: on the top shelf, a white book with a red patterned cover (category: book) is positioned at the center-left, next to a blue book and a black tablet. A hand from the viewer's perspective enters from the left to approach the white book. The hand is in an extended position, fingers open, approaching the top edge of the book's cover. The book has a vertical orientation, with its spine facing slightly toward the viewer.

Kinematics & Contact: In Frame 1, the hand is approaching the book with an open, relaxed grip. By Frame 2, the hand makes contact with the top and bottom edges of the book's cover, fingers curling into a precision pinch grip to lift the book. The contact points are the top (top edge of the cover) and bottom (bottom edge of the cover) of the white book. The hand then lifts the book upward and slightly backward (away from the shelf) during Frame 3, maintaining a secure precision pinch grip with the fingertips.

Final State: The white book with the red patterned cover has been lifted off the top shelf by the right hand. The hand now holds the book in a lifted position, away from the shelf, with the book oriented vertically and the precision pinch grip maintained on its cover edges. The grasp is still active, and the book is no longer in contact with the shelf.
7
1_1_1_000272
Action Summary: Right hand picks up the book (category: literature/reading material) from the shelf and places it on a different section of the shelf.  
Initial State: Frame 1 shows a black bookshelf with multiple items. On the upper shelf, a white book with a red patterned cover is visible. Other items include a green book, a blue book, and cardboard boxes. A right hand extends toward the white book, making contact with its cover in a precision grasp.  
Kinematics & Contact: From Frame 1 to Frame 2, the right hand grips the book’s cover (contact point: cover surface) with a precision grasp. The hand lifts the book upward and toward the center of the shelf. By Frame 3, the hand has released the book, which now rests on the shelf.  
Final State: The white book with the red cover is repositioned on the upper shelf, resting among other items (e.g., the green book, blue book, and a white box). The hand has released the book, and grasp is no longer maintained.  
Context: The action occurs in a static bookshelf environment with no tool use; only hand and object interaction. Secondary items (books, boxes) remain stationary and unaffected.
8
1_1_1_000310
Action Summary: Right hand picks up the green book from the shelf, moving it toward a vertical orientation.  
Initial State: At Frame 1, a green book with a white spine lies on a black shelf. The book is positioned between a blue book (lateral to it) and other items (distally, including a stack of books and a notebook). The right hand approaches the book from its lateral side, fingers extended to grasp the book’s base.  
Kinematics & Contact: The right hand transitions from a distal, open-palm approach to a power grasp, securing the book’s base. Grip type is a palmar grasp with distal-to-proximal finger closure. Contact is maintained as the hand lifts the book, transitioning it from a horizontal shelf position to a raised, vertical orientation.  
Final State: Frame 3 shows the green book held in the right hand, now upright (90° from the shelf). The book’s orientation has changed from horizontal (initial shelf state) to vertical, with the hand positioned laterally, supporting the book’s base. Other books on the shelf (blue book, stacked books) remain static. No secondary objects enter or exit the frame, and no tools are used.
9
1_1_1_000348
Action Summary: Right hand performs a grasp and lift to pick up a book from a black shelf; the target is a blue-green book (category: book), with the goal of moving it toward the upper shelf area.

Initial State: In Frame 1, a right hand approaches a blue-green book positioned upright on a black shelf. The book measures approximately 20 cm in height, has a rectangular shape with a green upper section and blue lower section, and rests on its base. The hand is in a proximal position relative to the book, approaching the base to initiate contact.

Kinematics & Contact: From Frame 1 to Frame 3, the hand maintains contact with the book’s base using a power grasp/palmar grasp grip. In Frame 2, the hand lifts the book, rotating it slightly in the hand while keeping contact with the base. No significant changes in grip type or contact points occur during this sequence; the hand applies consistent downward pressure at the base during lifting.

Final State: The book has been lifted off the shelf; its new position is near the upper portion of the shelf, among other books. The hand has released the book, and grasp is no longer maintained; the book is now stabilized in the hand during movement toward the target shelf area.
10
1_1_1_000388
Action Summary: Right hand picks up a book from the middle shelf, moves it toward the left, and releases it onto a lower shelf.  
Initial State: At Frame 1, a right hand is positioned near the top of a stack of books on the middle shelf; the top book is green, with a white book, a “C++ Primer” book, and another white book below it. The hand faces the top (green) book’s top surface, preparing to grasp.  
Kinematics & Contact: In Frame 1, the hand approaches the green book’s top surface, making initial contact. By Frame 2, the hand lifts the green book slightly after grasping it (grip type likely precision or power grasp, contact at the top surface). In Frame 3, the hand moves the green book leftward and places it onto a lower shelf, releasing it; grasp is maintained during lifting and released upon placement.  
Final State: The green book is now on the lower shelf, displaced from its original position on the middle shelf; the right hand has released the green book.  
Context: No tools are used. Secondary objects include a blue book standing on the middle shelf’s left and other books/papers on surrounding shelves, which remain stationary and unaffected by the action.
11
1_1_1_000426
Action Summary: Right hand performs a "pick and place" action, grasping a blue hardcover book from a lower shelf and placing it on an upper shelf.  

Initial State: In Frame 1, a right hand extends toward a blue hardcover book positioned upright on a black shelf. The book (small to medium size, rectangular, blue cover with text) is situated in the upper - right section of the shelf. The hand approaches from the left, fingers slightly curled, preparing for grasp.  

Kinematics & Contact: During Frame 1, the hand makes contact with the book’s top edge using a precision grasp (fingers around the cover, thumb on the side). In Frame 2, the hand lifts the book slightly, adjusting its grip for stability as it begins to move. The book remains gripped firmly (similar precision grasp) as the hand reorients it toward the upper shelf. In Frame 3, the hand moves the book horizontally toward the now - empty upper shelf, maintaining contact with the book’s base throughout the motion.  

Final State: The blue hardcover book is placed upright on the upper shelf, with the hand having released it. The grasp is no longer maintained, and the book stands vertically on the shelf.  

Context: No tools are used. The action occurs within a bookshelf environment; other books and items (e.g., a green - covered book, stacked books, a red - spined book) remain stationary. The hand’s movement is confined to the shelf area, focusing on repositioning the blue book.
12
1_1_1_000464
Action Summary: Place. The target object is a blue book titled "FOOD AND COOKING". The goal is to position it onto the upper shelf above the existing stack.  
Initial State: In Frame 1, a right hand (distal, lateral to the shelf) grasps the topmost blue book (category: book, size medium, rectangular, blue cover with text) in a power grasp, contact at the book’s top surface. The hand is extended toward the shelf.  
Kinematics & Contact: From Frame 1 to 2, the hand lifts the book upward and toward the shelf. Grip type remains a firm grasp, maintaining contact with the book’s top. Between Frame 2 and 3, the hand repositions the book laterally (toward the shelf’s center) while continuing the upward movement, still gripping the book.  
Final State: The book is now resting on the shelf, aligned atop the lower stack of books. The hand has released the book, so grasp is no longer maintained.  
Context: The action occurs on a dark-colored open shelf. Below the action area, additional books and items are visible. No tools or secondary objects directly interact with the main action, but the shelf’s structure (open compartments) frames the placement.
13
1_1_1_000502
Action Summary: Right hand grasps the blue book titled "FOOD AND COOKING", lifts it from the stack on the shelf, and moves it away from the original position.

Initial State: A blue book ("FOOD AND COOKING" by Harold McGee) rests atop a stack of books and notebooks on a black shelf. The right hand approaches the blue book from the right side, positioned to grasp it with a power grasp (fingers wrap around the book's top edge and side).

Kinematics & Contact: In Frame 1, the hand contacts the blue book’s top and lateral surface. From Frame 1 to Frame 3, the hand moves upward and leftward, lifting the book. The grip remains consistent, maintaining contact with the book’s top and side throughout the motion.

Final State: At Frame 3, the blue book is held in the right hand, removed from the original stack. The book’s orientation is unchanged, and grasp is maintained as the hand moves it away from the shelf.
14
1_1_1_000540
Action Summary: Right hand picks up the blue book (category: book) from the top shelf, with the goal of moving the book.  
Initial State: Frame 1 shows a right hand approaching a blue book on a black bookshelf. The hand is positioned laterally in front of the shelf, fingers extended to grasp the book’s cover. Other objects on the shelf include a green book, a “C++ Primer” book, and stacked books below. The hand’s orientation is toward the book, with a tentative grip forming around the cover.  
Kinematics & Contact: From Frame 1 to Frame 2, the hand moves closer to the blue book, adjusting grip to secure it (precision pinch grasp). Contact is maintained on the book’s cover throughout, with no significant changes in force or grip type. Frame 3 shows the hand lifting the book away from the shelf, with the book’s cover oriented upward and the grip still firm on the cover.  
Final State: At Frame 3, the blue book has been lifted off the shelf, held in the right hand. The book’s position is now above the shelf, with the grip maintained on the cover.  
Context: The action occurs on a black bookshelf containing multiple books (green, technical books like “C++ Primer,” and stacked books below). The environment remains static, with no other objects or tools interacting with the hand or the book during this sequence.
15
1_1_1_000578
Action Summary: Right hand picks up the book (category: book) from the shelf, lifting it toward a more accessible position.  
Initial State: At Frame 1, a right hand is distal from a blue book on a black shelf; the shelf holds multiple books (green, white with text, stacked books including “C++ Primer”). The hand is positioned to grasp the blue book.  
Kinematics & Contact: From Frame 1 to 2, the right hand transitions from approaching the blue book to gripping its spine/edge with a precision pinch. Contact is maintained as the hand lifts the book.  
Final State: At Frame 3, the blue book has been lifted, with the hand now holding it away from the shelf. Grasp is maintained, and the book’s position/orientation has changed from resting on the shelf to being held.  
Context: No tools; the action focuses on manipulating the blue book on a shelf with other books.
16
1_1_1_000616
Action Summary: Right hand picks up the green book from the shelf using a power grasp, lifting it toward a more open area of the shelf.  
Initial State: At Frame 1, a black bookshelf holds multiple books; the green book (with a white spine and green cover) is the topmost item in a small stack. A right hand extends toward the green book, palm facing the object. The hand is positioned laterally adjacent to the book stack.  
Kinematics & Contact: In Frame 2, the right hand executes a power grasp around the base of the green book (where the cover meets the spine), fingers curling around the edge. The book remains stationary as the hand applies force to lift it. In Frame 3, the hand lifts the green book, transitioning from contact with the base to gripping the bottom edge as it moves away from the shelf.  
Final State: The green book is now held by the right hand, lifted away from the shelf with a maintained power grasp; the book’s cover faces forward, and it is oriented upright.
17
1_1_1_000654
Action Summary: Right hand picks up the green book (category) from the left section of the black shelf and places it onto the shelf.  
Initial State: In Frame 1, a right hand grasps a green book (size: small to medium, rectangular, hardcover) located in the left vertical section of the black shelf. The book is oriented vertically, with its cover facing right. The hand is in a precision grasp, fingers wrapped around the book’s cover, positioned toward the top-left of the book.  
Kinematics & Contact: From Frame 1 to Frame 3, the hand moves the green book horizontally toward the center of the shelf while maintaining a firm precision grip on the cover. The contact point remains the book’s cover throughout; no grip changes occur, and the hand remains in a single, steady grasp.  
Final State: At Frame 3, the green book is placed on top of other books in the center of the shelf, oriented vertically with its cover facing upward. The hand has released the book, so grasp is deactivated.  
Context: The black shelf contains other books (e.g., a white book with red geometric patterns, a blue book, and a “C++” book) in the central and right sections. The action involves only the hand and the green book, with no tools or secondary environmental changes affecting the action.
18
1_1_1_000692
Action Summary: Right hand grasps the green book (category: book) from the upper shelf and lifts it away from the shelf.  
Initial State: Right hand approaches the green book (shape: rectangular, size: medium, color: green with white cover) positioned on the upper shelf; the hand is in a power grasp near the book’s lateral face.  
Kinematics & Contact: In Frame 1, the right hand establishes a grip on the green book’s base using a palmar grasp. From Frame 2 to 3, the hand lifts the book vertically, maintaining contact on the base throughout; no significant grip or contact changes.  
Final State: The green book is now held by the right hand, positioned away from the upper shelf, with the hand in a relaxed grasp (grip maintained).  
Context: The action occurs on a black bookshelf with other books (e.g., “C++ Primer”, a white book with red text, a blue book) and notebooks visible; no tools or secondary objects directly affect the action.
19
1_1_1_000730
Action Summary: Right hand picks up the green book from the top shelf, moving it toward the left.  
Initial State: Frame 1 shows a green book (category: book, size: medium, shape: rectangular with rounded corners, orientation: cover facing upward, position: top shelf left section) on the top shelf; a right hand (forearm visible, arm extended toward the book) is positioned near the book, preparing to grasp it.  
Kinematics & Contact: The right hand moves from the lower left area of the frame toward the green book, making contact with the book’s cover and spine using a precision grasp (fingers wrapping around the edges). In Frame 2, the hand lifts the book slightly, maintaining contact with the cover. In Frame 3, the hand continues lifting the book away from the shelf, with the contact point remaining on the book’s cover.  
Final State: The green book is now being held by the right hand, positioned above the shelf (away from its initial resting place), with the grasp maintained on the cover.  
Context: No tools or significant environmental changes; the action focuses solely on manipulating the green book.
20
1_1_1_000768
Action Summary: Pick, target a green book (category: book), goal position it on the shelf among other books.

Initial State: Frame 1 shows a right hand (proximal to the viewer, lateral on the shelf) making contact with the top edge of a green-covered book on the shelf; the book is upright with visible white pages, and the hand is positioned to grasp it.

Kinematics & Contact: The hand transitions from a precision grasp (fingers curling under the top edge of the green book) in Frame 1 to lifting the book in Frame 2, adjusting grip mid-air. In Frame 3, the hand releases the green book, which settles onto the shelf among other books.

Final State: The green book is now resting on the shelf, displaced from its initial position to be among other stacked books, and the hand has moved away from the object.
21
1_1_1_000806
Action Summary: Right hand picks up the top book (category: book) from the black shelf, grasping it with a palmar grasp, and lifts it toward an upper position outside the frame.  

Initial State: A black shelf holds multiple books: a green-covered book sits atop a stack of books (including one with “C++” visible), and a blue book stands upright on the right. The right hand is extended toward the green - covered book, fingers adjusting to grasp its top edge.  

Kinematics & Contact: In Frame 1, the right hand approaches the green - covered book’s top edge. By Frame 2, the hand establishes a palmar grasp, with fingertips and the base of the thumb contacting the book’s cover. In Frame 3, the hand lifts the book upward, moving it away from the shelf, while maintaining a secure grip via the same contact points.  

Final State: The green - covered book has been successfully picked up and is now held in the right hand, removed from the shelf. The book’s orientation (cover facing outward, edges aligned) and the grasp type (palmar grip) remain consistent from the start to finish of the action.  

Context: No tools are used. The action occurs in a static environment focused on the shelf; no secondary objects enter or leave the immediate action space.
22
1_1_1_000844
Action Summary: Right hand picks up the green book from the black shelf, lifting it toward the upper shelf (or removing it from the shelf).  
Initial State: At Frame 1, the right hand is extended toward a black shelf; the hand is positioned to grasp a green book resting atop other books (including a red “C++” book and a white book with red patterns). The hand is in a power grasp, making contact with the green book’s cover and side.  
Kinematics & Contact: The hand applies a power grasp with positive force, gripping the green book’s cover and lateral side. The trajectory moves the hand upward and slightly leftward, lifting the book away from the shelf. In Frame 2, the book is lifted off the shelf, and in Frame 3, the hand continues to lift the book toward the upper shelf area, with the book’s orientation altered from its initial vertical stack position.  
Final State: The green book is being held by the right hand, lifted upward and to the left, away from the black shelf it was on. The grasp is maintained during this action, with the book’s cover and side firmly gripped.  
Context: The black shelf contains multiple books (including a red “C++ PRIMER” book and a white book with red geometric patterns beneath the green book). To the right of the shelf, a blue book stands upright. Below the shelf, additional books are visible on a lower shelf. A beige mug sits on a wooden surface (possibly a nightstand) to the right of the frame.
23
1_1_1_000882
Action Summary: Right hand performs a pick action on a book, lifting it from the top shelf.  
Initial State: In Frame 1, a black bookshelf holds multiple objects: a white book with red patterns, a “C++” book, a black rectangular object (possibly a tablet), and a white book, all stacked on the top shelf. A right hand approaches the green-covered book positioned on the right side of the shelf.  
Kinematics & Contact: The hand moves from the lower front region of the shelf toward the green book. Contact is established with the book’s cover via a precision grasp (or similar grip) as the hand secures the object. No significant changes in grip, contact points, or force are observed across frames.  
Final State: By Frame 3, the book has been lifted off the shelf; it is now held in the right hand, displaced from its original position on the top shelf. The grasp is maintained during the action.  
Context: The environment consists of a black bookshelf with books and a partial view of a bed with striped bedding to the right. No tools or secondary objects directly interact with the action beyond the book being picked.
24
1_1_1_000920
Action Summary: Right hand grasps the book with a red patterned cover and lifts it toward the upper shelf.

Initial State: At Frame 1, a black bookshelf holds multiple books. The target book (white cover with red geometric pattern, spine text partially visible as "计算机科学导论") is positioned on the upper shelf, angled slightly toward the right. A right hand extends from the lower left, fingers curling for a precision grasp over the book’s cover, with the thumb on the left edge and thumb tip pressing against the upper cover.

Kinematics & Contact: From Frame 1 to Frame 2, the right hand moves upward and forward, fingers closing around the book’s cover (contact points: left edge, upper surface, and back cover). Grip is a precision pinch (thumb and index finger), with the book’s base (black spine) lifting slightly. By Frame 3, the hand repositions the book horizontally, then releases it onto the shelf as the fingersMy understanding is that you want me to act as a human expert annotating this video for robotics. I need to be very precise, using spatial terms and quantifying where possible. Let's go through each step.

First, analyze the action. The person is picking up a book and placing it. So the action primitive is "pick" or "place"? Wait, the sequence shows picking and then placing. Let's check the frames. The hand reaches for the book, grasps it, lifts, then places it.

Initial State: Frame 1 shows the shelf with books. The target book is on the upper shelf, with a red pattern on the cover. The hand is approaching from the left (lower left of the shelf).

Kinematics: The hand trajectory is from lower left, moving up and forward to grasp the book. Contact is on the book's cover (specifically the upper left corner and the red pattern area). Grip is a precision grasp with the thumb and fingers.

Final State: After Frame 3, the book is placed on the shelf, now in a new position (horizontal, resting on other books below). The hand is moving away.

Context: The shelf is black, made of what looks like particle board. Other books are visible (green, blue, white, etc.), and there's a wooden cabinet on the right. No tools are used, just the hand.

Now, structure the output as specified.

Action Summary: Right hand picks up the book with a red patterned cover from the upper shelf and places it on the lower shelf.

Initial State: At Frame 1, a black bookshelf contains various books, including a prominent book with a white cover and red geometric pattern (spine text partially visible as “计算机科学导论”) positioned on the upper shelf. A right hand extends from the lower left, fingers curling into a precision pinch grasp over the book’s upper cover (contact points: left edge, upper surface, and back cover).

Kinematics & Contact: Between Frame 1 and Frame 2, the right hand moves upward and forward, closing around the book’s cover using a precision pinch (thumb on left edge, index finger on upper cover, other fingers curled inward). The book’s base (black spine) lifts slightly as the hand lifts it. In Frame 3, the hand repositions the book horizontally, then releases it onto the lower shelf, with the book now resting on other books beneath it.

Final State: By Frame 3, the book with the red patterned cover has been placed on the lower shelf, resting on top of other vertically oriented books. The hand has moved away, and the grasp is released.

Context: The scene occurs in an indoor environment with a black bookshelf as the central object. Additional books (green, blue, white, etc.) are stored on both shelves, and a wooden cabinet is visible to the right. No tools or external objects directly interact with the action beyond the hand and shelf.
25
1_1_1_000958
Action Summary: Right hand grasps the book (category: book) from the middle shelf and lifts it toward the upper shelf area.  
Initial State: In Frame 1, the right hand is extended toward a book on the middle shelf. The book has a white cover with red/black patterned edges and is titled “Python”. Other books on the same shelf include a green hardcover, a blue hardcover, and a black-covered book. Lower shelves hold additional books and items. The hand is positioned to grasp the book using a precision or power grasp, with contact likely on the book’s base or cover.  
Kinematics & Contact: From Frame 1 to 2, the hand moves closer to the book, adjusting its grip (possibly transitioning from a loose grasp to a firmer one). In Frame 2, the hand lifts the book, with the grip maintained on the book’s base/cover as it begins to detach from the shelf. No significant changes in grip type or contact points occur in the observed frames.  
Final State: By Frame 3, the book is no longer in contact with the shelf and is held aloft by the right hand, positioned above the original shelf level as it is being moved upward.  
Context: The action takes place in a room with a dark-colored bookshelf. No tools are used; the manipulation is entirely manual. The hand moves from a position near the shelf to an upward trajectory, with no other objects entering or leaving the immediate manipulation area.
26
1_1_1_001001
Action Summary: Right hand grasps the book from the left side of the shelf and lifts it.  
Initial State (Frame 1): A right hand is extended toward a white book with red geometric patterns on its cover, positioned upright on the left side of a black shelf. The book is hard - covered, rectangular, and appears to have text and red decorative designs. The hand adopts a power grasp, with fingers curling around the book’s top edge and palm supporting from the back.  

Kinematics & Contact (Frames 1–3): From Frame 1 to 2, the hand moves closer to the book, adjusting its grip. In Frame 2, fingers apply pressure to the book’s spine/top edge, while the palm supports from the rear. By Frame 3, the hand lifts the book upward and slightly backward, transitioning from the grasping position to the lifting phase. A power grasp (fingers and palm) maintains contact, with the key contact points being the book’s top edge and spine.  

Final State: The book is no longer in its original position on the shelf; it is lifted into the hand, which is now positioned above and behind the shelf. The grasp is still maintained, with the book held ready for placement.  

Context: Other objects on the shelf (a tablet, a green folder, and additional books) remain static during this pick action. No tools or secondary environmental changes directly impact the book - picking process.
27
1_1_1_001047
Action Summary: Right hand picks up the "C++" book from the middle shelf and lifts it toward the upper shelf.  
Initial State: In Frame 1, a right hand is positioned near the "C++" book (category: book, white cover with red decorative elements, spine oriented horizontally) on the middle shelf. The hand is distal to the book, approaching from the right side of the frame.  
Kinematics & Contact: From Frame 1 to Frame 2, the hand moves toward the book, making contact via a precision grip around the book’s spine and upper cover. In Frame 3, the hand retracts while maintaining grip, lifting the book upward. The trajectory is vertical, moving the book from the middle shelf toward the upper shelf.  
Final State: The "C++" book is now placed on the upper shelf, positioned vertically with its spine facing outward, similar to its initial orientation but relocated to the upper shelf. Grasp is maintained throughout the action and released upon placement.  
Context: No tools are used; the action involves a single hand manipulating a book within a bookshelf environment, with no secondary objects entering or leaving the frame that directly affect the action.
28
1_1_1_001085
Action Summary: Right hand performs a pick-and-place action, grasping a book with a white cover from the upper shelf and moving it toward the upper right of the frame.

Initial State (Frame 1): A dark-colored bookshelf is visible, with a book featuring a white cover and partial text (possibly Chinese characters) positioned on the shelf. The right hand (proximal to the camera) is extended, fingers splayed, approaching the book’s cover. The book is upright, rectangular, and appears to be of medium size (height ~12–15 inches, width ~5–7 inches, depth ~1 inch). The hand’s grip orientation is external, preparing for grasp.

Kinematics & Contact (Frames 1–3): In Frame 1, the hand makes contact with the book’s cover using a power grasp (fingers wrapped around the top edge). The contact point is the upper surface of the book. In Frame 2, the hand lifts the book slightly, maintaining contact along the top and side edges. The grip adjusts minimally to secure the book as it begins to rise. In Frame 3, the hand has moved the book away from the original shelf position, holding it with a consistent power grasp; the book’s orientation remains upright, and contact is maintained on the top cover.

Final State (Frame 3): The book (white cover, partial text) is now displaced from its original shelf location, held by the right hand toward the upper right area of the frame. The grasp remains active, and the book’s orientation is preserved (upright, text-side facing forward relative to the hand).
29
1_1_1_001123
Action Summary: Right hand picks up the C++ book from the top shelf, lifting it toward the lower shelf area.  
Initial State: At Frame 1, a white book with a gray and red cover (labeled “C++”) sits on the top shelf, positioned next to a green book (right) and a black tablet-like object (left). A right hand approaches from the right, fingers extended to grasp the book’s cover.  
Kinematics & Contact: From Frame 1 to 2, the right hand moves laterally toward the book, settling into a precision grasp (fingers curled around the book’s top edge and back cover). Contact points are the book’s top-right corner and upper back. Force applied is steady as the hand begin to lift the book.  
Final State: By Frame 3, the C++ book has been lifted and is positioned above the lower shelf (where other books are stored). The hand maintains a grasp on the book, actively repositioning it.
30
1_1_1_001161
**Action Summary:** The action is a "pick and place" operation where a hand grasps a book (C++ programming book) from the upper shelf and reposition it to the left side of the same shelf, next to the book titled "MONET."  

**Initial State (Frame 1):** A right hand, in a precision grip, approaches a book with a red spine label ("C++") from the right side of the upper shelf. The book rests on top of two stacked white books ("MONET"). Nearby objects include a green book, a blue book, and other vertically stacked books. The hand is distal to the book, positioned to grasp its spine.  

**Kinematics & Contact (Frames 1–3):**  
- **Frame 1:** The hand initiates contact with the book’s spine using a precision pinch grip. The book is oriented horizontally, with its front cover facing outward.  
- **Frame 2:** The hand lifts the book vertically, moving it upward and slightly leftward. The grip remains a precision pinch on the spine, and the book’s orientation changes to a vertical stack position.  
- **Frame 3:** The hand lowers the book onto the shelf, placing it next to the "MONET" book. The grip transitions to a plant/grasp pattern, then releases the book as it settles into its new position.  

**Final State:** The book ("C++") is stationary on the upper shelf, positioned upright and adjacent to the "MONET" book. The hand has retracted, and the book’s orientation is now vertical. The surrounding books remain unchanged.  

**Context:** No tools are used. Secondary objects (e.g., green book, blue book) remain static. The action consists of a single bimanual grasp-and-place sequence, with no significant environmental changes beyond the book’s repositioning.
31
1_1_1_001199
Action Summary: Right hand grasps the "MONET" book (category: book) and lifts it toward the upper shelf.

Initial State: Frame 1 shows a black bookshelf with multiple books on the middle shelf. A white-covered book titled "MONET" is positioned on top of another book, slightly angled. The right hand (from the bottom - left of the frame) is moving toward the "MONET" book, which is approximately 15 cm wide and 5 cm thick, with a smooth finish.

Kinematics & Contact: In Frame 1, the hand approaches the "MONET" book from the distal side. Frame 2 shows the hand making contact with the top surface of the "MONET" book using a precision grip, with fingers curled around the book's top edge. In Frame 3, the hand lifts the "MONET" book upward, moving it away from the middle shelf toward the upper shelf area.

Final State: The "MONET" book is now held by the right hand, lifted off the middle shelf. The grasp is maintained, and the book is oriented with its cover facing forward. The original position of the "MONET" book on the middle shelf is now empty.
32
1_1_1_001237
Action Summary: Right hand performs a pick action, grasping the stack of books (with the top book titled "MONET") from the black bookshelf.  
Initial State: Frame 1 shows a right hand in a proximal position, making contact with the top book (titled "MONET") of a small stack on a black bookshelf. The hand uses a precision grasp (thumb on the book’s top edge, fingers lifting the base), with the "MONET" book being the target. Other objects on the shelf include a C++ book, a Chinese-language book, and a green book. The hand’s configuration is static, preparing to lift.  
Kinematics & Contact: From Frame 1 to 2, the hand applies a precision grasp to the "MONET" book, maintaining contact along its top edge and base. In Frame 3, the hand begins to lift the book stack, transitioning from contact on the top (grasping the cover) to lifting the entire stack away from the shelf, with the grip persisting on the top book.  
Final State: The stack of books (including "MONET") is now elevated above the shelf, with the hand still grasping the top book. The books are no longer in contact with the shelf, and the grasp is maintained throughout the action sequence.  
Context: The action occurs in a room with a black bookshelf holding various books; no tools or secondary objects directly interact with the hand or books beyond the shelf itself.