arXiv:2609.09736v1 Announce Type: new Abstract: Video Temporal Grounding (VTG) localizes the video segment that matches a natural-language query. Many queries describe an action performed by a partic...
arXiv:2606. 29613v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) architectures have recently been extended with role-based mechanisms for interpretability.
arXiv:2504.10079v5 Announce Type: replace Abstract: Few-shot action recognition (FSAR) aims to recognize novel action categories with few exemplars. Existing methods typically learn frame-level repre...
The paper introduces the Identity-Aware Human-Object Interaction Motion Captioning task, which requires captions to include both the subject’s identity and the interaction motion, e.g., "Sub_ID lifts the chair" instead of a generic description. It proposes ID‑HOINet, a model that learns from multi‑view videos using a Multi‑View Identity‑Motion Learning Module and a Two‑Stage Caption Rewriting Strategy to generate identity‑aware captions. Experiments show that ID‑HOINet achieves state‑of‑the‑art performance on the BEHAVE and InterCap datasets.
arXiv:2608. 10765v1 Announce Type: new Abstract: Recognizing human behavior across levels of abstraction, from atomic actions to long-horizon intentions, requires data annotated along a semantic hierarchy.
CogCanvas is a new benchmark for multi-subject reference-based image generation, featuring 1,952 curated reference images of 100 celebrities, 115 objects/fashion items, and 29 real-world backgrounds. It generates 1,361 compositional prompts with 2–5 people, using a pipeline that includes DINOv2 deduplication, aesthetic filtering, and automated graph derivation for interaction and positioning. The benchmark evaluates three tasks—reference-based multi-human-object generation, text-to-image compositional generation, and reference retrieval—under a six-axis protocol, and introduces BG‑Sim and Attr‑VQA metrics to assess background fidelity and attribute binding.
arXiv:2606. 32017v1 Announce Type: cross Abstract: Agentic reinforcement learning requires assigning credit to environment-facing actions such as searches, clicks, edits, navigation commands, and object interactions.
arXiv:2607. 13056v1 Announce Type: cross Abstract: Current vision-language-action (VLA) benchmarks primarily evaluate isolated manipulation skills while leaving human-robot interaction structure largely unmodeled.
ActionPiece rethinks how actions are tokenized for autoregressive vision‑language‑action models by introducing physical rank consistency (PRC) to evaluate relational fidelity of reconstructed actions. The method jointly supervises representation learning and quantization to preserve local physical distance rankings, improving both PRC and policy success. Experiments on LIBERO, LIBERO‑Plus, SimplerEnv, and VLA‑Arena show significant gains over baseline tokenizers.
arXiv:2608. 20127v1 Announce Type: new Abstract: Video Temporal Grounding (VTG) faces significant challenges when natural language queries must distinguish between multiple events involving visually similar entities, particularly when relying on fine-grained visual attributes that are difficult to describe accurately in words alone.
arXiv:2609.40219v1 Announce Type: cross Abstract: World Action Models (WAMs) couple visual dynamics prediction with action generation, yet they do not explicitly support the reuse of action experienc...
RotVLA introduces a Vision‑Language‑Action framework that replaces discrete latent action encoding with a continuous rotational latent action representation on the group SO(n). This design provides continuity, compositionality, and structured geometry that better capture real‑world action dynamics, and a triplet frame learning scheme enforces meaningful temporal dynamics while preventing degeneration. Trained with 1.7 B parameters on large cross‑embodiment datasets, RotVLA achieves state‑of‑the‑art performance on LIBERO and RoboTwin2.0 benchmarks and shows strong real‑world manipulation results.