animation/tools/mixamo_retarget.py

# Mixamo FBX → Quaternius-skeleton GLB batch retargeter (Blender headless).
#
# Sibling of tools/kevin_retarget.py — same world-rotation-delta method, adapted for the
# Mixamo rig ("mixamorig:*" bones, prefix auto-detected). Both rigs rest in a T-pose, so
# for every mapped bone the world rotation delta from rest transfers 1:1:
#     q_target_world(f) = (q_src_world(f) · q_src_rest⁻¹) · q_target_rest
# Pelvis translation transfers as a WORLD-space delta from rest (rotated through the
# frame-alignment quaternion C, scaled by the hips→head size ratio) — unlike Kevin's
# raw-fcurve read, because Mixamo roots the armature at the hips with no demo-scene
# offset bone, and world deltas survive whatever local axes the importer picked.
# Unmapped target bones (leaf/ball bones etc.) ride their parent rigidly at rest offset.
#
# Usage:
#   blender --background --python tools/mixamo_retarget.py -- \
#       --src "<dir-or-fbx>[,<dir-or-fbx>…]" --out <out.glb> [--limit N] [--target <gltf>]

import bpy, sys, os, re
from mathutils import Matrix, Quaternion, Vector

REPO_ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
DEFAULT_TARGET_GLTF = os.path.join(
    REPO_ROOT, "assets", "quaternius", "base-characters",
    "Universal Base Characters[Standard]", "Base Characters", "Godot - UE",
    "Superhero_Male_FullBody.gltf")

# Mixamo (prefix-stripped) → Quaternius UE names. Fingers map 1:1 (both rigs 3-segment).
BONE_MAP = {
    "Hips": "pelvis",
    "Spine": "spine_01",
    "Spine1": "spine_02",
    "Spine2": "spine_03",
    "Neck": "neck_01",
    "Head": "Head",
}
for S, T in (("Left", "l"), ("Right", "r")):
    BONE_MAP[f"{S}Shoulder"] = f"clavicle_{T}"
    BONE_MAP[f"{S}Arm"] = f"upperarm_{T}"
    BONE_MAP[f"{S}ForeArm"] = f"lowerarm_{T}"
    BONE_MAP[f"{S}Hand"] = f"hand_{T}"
    BONE_MAP[f"{S}UpLeg"] = f"thigh_{T}"
    BONE_MAP[f"{S}Leg"] = f"calf_{T}"
    BONE_MAP[f"{S}Foot"] = f"foot_{T}"
    BONE_MAP[f"{S}ToeBase"] = f"ball_{T}"
    for i in (1, 2, 3):
        BONE_MAP[f"{S}HandIndex{i}"] = f"index_0{i}_{T}"
        BONE_MAP[f"{S}HandMiddle{i}"] = f"middle_0{i}_{T}"
        BONE_MAP[f"{S}HandRing{i}"] = f"ring_0{i}_{T}"
        BONE_MAP[f"{S}HandPinky{i}"] = f"pinky_0{i}_{T}"
        BONE_MAP[f"{S}HandThumb{i}"] = f"thumb_0{i}_{T}"


def detect_prefix(arm):
    """Mixamo bones are usually 'mixamorig:Hips' but exports vary ('mixamorig1:', none)."""
    for b in arm.data.bones:
        if b.name.endswith("Hips"):
            return b.name[: -len("Hips")]
    raise RuntimeError("no *Hips bone found — not a Mixamo rig?")


def apply_object_transform(obj):
    """Bake importer object-level rotation/scale into the armature data (meters, Z-up)."""
    bpy.ops.object.select_all(action="DESELECT")
    obj.select_set(True)
    bpy.context.view_layer.objects.active = obj
    bpy.ops.object.transform_apply(location=True, rotation=True, scale=True)
    bpy.context.view_layer.update()


def clip_name(path):
    base = os.path.splitext(os.path.basename(path))[0]
    base = base.replace(" - ", "_").replace(" ", "")
    return re.sub(r"[^A-Za-z0-9_\-]", "", base)


def hierarchy_order(arm):
    out, stack = [], [b for b in arm.data.bones if b.parent is None]
    while stack:
        b = stack.pop(0)
        out.append(b)
        stack.extend(b.children)
    return out


def char_frame(rest, hips, head, upper_arm_l):
    """Orthonormal character frame from the rest pose: columns = (left-arm axis, hips→head up,
    their cross) — aligns the two rigs whatever world orientation the importer left them in."""
    u = (rest[head].translation - rest[hips].translation).normalized()
    l = (rest[upper_arm_l].to_quaternion() @ Vector((0, 1, 0))).normalized()  # bone Y = along bone
    f = l.cross(u).normalized()
    l = u.cross(f).normalized()
    m = Matrix((l, u, f)).transposed()
    return m.to_quaternion().normalized()


def retarget_one(src_arm, tgt_arm, action, name, scn, prefix, unit):
    f0, f1 = int(action.frame_range[0]), int(action.frame_range[1])
    if src_arm.animation_data is None:
        src_arm.animation_data_create()
    src_arm.animation_data.action = action
    # Blender 5 slotted actions: bind the first slot so the action actually evaluates.
    if getattr(action, "slots", None) and src_arm.animation_data.action_slot is None:
        src_arm.animation_data.action_slot = action.slots[0]

    inv_map = {v: prefix + k for k, v in BONE_MAP.items()}

    tgt_rest = {b.name: (tgt_arm.matrix_world @ b.matrix_local) for b in tgt_arm.data.bones}
    src_rest = {b.name: (src_arm.matrix_world @ b.matrix_local) for b in src_arm.data.bones}
    src_rest_q = {n: m.to_quaternion().normalized() for n, m in src_rest.items()}
    tgt_rest_q = {n: m.to_quaternion().normalized() for n, m in tgt_rest.items()}

    C = (char_frame(tgt_rest, "pelvis", "Head", "upperarm_l")
         @ char_frame(src_rest, prefix + "Hips", prefix + "Head", prefix + "LeftArm").inverted()).normalized()
    Cinv = C.inverted()

    # Uniform size ratio from hips→head world distance (unit/scale agnostic).
    h_src = (src_rest[prefix + "Head"].translation - src_rest[prefix + "Hips"].translation).length
    h_tgt = (tgt_rest["Head"].translation - tgt_rest["pelvis"].translation).length
    tscale = h_tgt / h_src if h_src > 1e-5 else 1.0

    new_act = bpy.data.actions.new(name)
    if tgt_arm.animation_data is None:
        tgt_arm.animation_data_create()
    tgt_arm.animation_data.action = new_act

    order = hierarchy_order(tgt_arm)
    tw_inv = tgt_arm.matrix_world.inverted()
    for pb in tgt_arm.pose.bones:
        pb.rotation_mode = "QUATERNION"

    src_hips_rest_t = src_rest[prefix + "Hips"].translation

    for f in range(f0, f1 + 1):
        scn.frame_set(f)
        dg = bpy.context.evaluated_depsgraph_get()
        se = src_arm.evaluated_get(dg)
        spose = {pb.name: (src_arm.matrix_world @ pb.matrix) for pb in se.pose.bones}

        tgt_world = {}
        for tb in order:
            nt = tb.name
            rest_t = tgt_rest[nt]
            ns = inv_map.get(nt)
            if ns is None or ns not in spose:
                if tb.parent is None:
                    tgt_world[nt] = rest_t
                else:
                    rel = tgt_rest[tb.parent.name].inverted() @ rest_t
                    tgt_world[nt] = tgt_world[tb.parent.name] @ rel
                continue
            q_delta = (spose[ns].to_quaternion().normalized() @ src_rest_q[ns].inverted()).normalized()
            q_delta = (C @ q_delta @ Cinv).normalized()   # express the delta in TARGET world axes
            q_world = (q_delta @ tgt_rest_q[nt]).normalized()
            if nt == "pelvis":
                # World-space hips delta from rest → target world, size-scaled. Carries the
                # dance's floor work (crouches, drops) without caring about local axes.
                # × unit: transform_apply bakes the importer's cm→m scale into bone REST data,
                # but the hips location FCURVES stay in cm — the raw world delta is 100× real.
                d = C @ (spose[ns].translation - src_hips_rest_t) * tscale * unit
                t_world = rest_t.translation + d
            else:
                rel = tgt_rest[tb.parent.name].inverted() @ rest_t
                t_world = (tgt_world[tb.parent.name] @ rel).translation
            tgt_world[nt] = Matrix.LocRotScale(t_world, q_world, Vector((1, 1, 1)))

        for tb in order:
            pb = tgt_arm.pose.bones[tb.name]
            m_arm = tw_inv @ tgt_world[tb.name]
            if tb.parent:
                m_parent_arm = tw_inv @ tgt_world[tb.parent.name]
                rel_rest = tb.parent.matrix_local.inverted() @ tb.matrix_local
                basis = rel_rest.inverted() @ (m_parent_arm.inverted() @ m_arm)
            else:
                basis = tb.matrix_local.inverted() @ m_arm
            pb.matrix_basis = basis
            pb.keyframe_insert("rotation_quaternion", frame=f)
            if tb.name == "pelvis" or tb.parent is None:
                pb.keyframe_insert("location", frame=f)

    tgt_arm.animation_data.action = None
    return new_act


def main():
    argv = sys.argv[sys.argv.index("--") + 1:] if "--" in sys.argv else []
    args = dict(zip(argv[::2], argv[1::2]))
    srcs, out = args["--src"].split(","), args["--out"]
    limit = int(args.get("--limit", 0))
    target = args.get("--target", DEFAULT_TARGET_GLTF)

    fbx_files = []
    for s in srcs:
        s = s.strip()
        if os.path.isdir(s):
            for root, _, files in os.walk(s):
                fbx_files += [os.path.join(root, x) for x in sorted(files) if x.lower().endswith(".fbx")]
        else:
            fbx_files.append(s)
    if limit:
        fbx_files = fbx_files[:limit]
    print(f"[mixamo] {len(fbx_files)} clips → {out}")

    bpy.ops.object.select_all(action="SELECT")
    bpy.ops.object.delete()
    scn = bpy.context.scene
    scn.render.fps = 30

    bpy.ops.import_scene.gltf(filepath=target)
    tgt_arm = next(o for o in bpy.data.objects if o.type == "ARMATURE")
    tgt_arm.name = "Armature"
    # Drop meshes — animations-only GLB (the game binds clips to its own body).
    for o in [o for o in bpy.data.objects if o.type == "MESH"]:
        bpy.data.objects.remove(o, do_unlink=True)
    bpy.context.view_layer.update()
    apply_object_transform(tgt_arm)

    done = []
    for i, f in enumerate(fbx_files):
        name = clip_name(f)
        pre = set(bpy.data.objects)
        pre_actions = set(bpy.data.actions)
        bpy.ops.import_scene.fbx(filepath=f)
        new_objs = [o for o in bpy.data.objects if o not in pre]
        src_arm = next((o for o in new_objs if o.type == "ARMATURE"), None)
        new_acts = [a for a in bpy.data.actions if a not in pre_actions]
        if src_arm is None or not new_acts:
            print(f"[mixamo] SKIP (no armature/action): {f}")
        else:
            bpy.context.view_layer.update()
            prefix = detect_prefix(src_arm)
            # Importer unit scale (cm→m 0.01) — capture BEFORE transform_apply wipes it;
            # the hips translation math needs it (fcurves stay in source units).
            unit = src_arm.scale.x
            apply_object_transform(src_arm)
            act = retarget_one(src_arm, tgt_arm, new_acts[0], name, scn, prefix, unit)
            done.append(act)
            print(f"[mixamo] {i + 1}/{len(fbx_files)} {name} ({int(act.frame_range[1])}f, prefix='{prefix}')")
        for o in new_objs:
            bpy.data.objects.remove(o, do_unlink=True)
        for a in new_acts:
            bpy.data.actions.remove(a)

    # Stash every baked action on an NLA track so the glTF exporter emits it as an animation.
    for act in done:
        tr = tgt_arm.animation_data.nla_tracks.new()
        tr.name = act.name
        tr.strips.new(act.name, 1, act)
    bpy.ops.object.select_all(action="DESELECT")
    tgt_arm.select_set(True)
    bpy.context.view_layer.objects.active = tgt_arm
    bpy.ops.export_scene.gltf(
        filepath=out, use_selection=True,
        export_animations=True, export_animation_mode="NLA_TRACKS",
        export_force_sampling=True, export_optimize_animation_size=False)
    print(f"[mixamo] EXPORTED {len(done)} clips → {out}")


main()