关系浏览

筛选上下游、竞争、合作、引用及事件演进关系。

REFERENCES

scenic-area interactive kiosk→Public-Space Gesture Interaction

可信度 98%

The paper states the analysis is based on '8 engineering repair records from a scenic-area interactive kiosk project' and that Public-Space Gesture Interaction is 'deployed systems' in contexts including 'scenic kiosks'; the kiosk is a concrete deployment instance of the broader Public-Space Gesture Interaction system.

REFERENCES

FlowGuard→cross-modal consistency

可信度 98%

FlowGuard detects harmful inputs by monitoring internal multimodal consistency

REFERENCES

FlowGuard→Partial Information Decomposition

可信度 95%

FlowGuard derives FlowVectors inspired by Partial Information Decomposition

COOPERATES_WITH

Transferable Latency Prediction for Fast LLM Screening on Heterogeneous Edge Devices→NVIDIA

可信度 95%

The framework is validated on Jetson Nano (NVIDIA) and an RTX 3090-class GPU (NVIDIA), indicating integration and testing with NVIDIA hardware platforms.

REFERENCES

FlowEvo→GSM8K

可信度 95%

Experiments on benchmarks spanning interactive environments (ALFWorld) and code/math generation (HumanEval, GSM8K)...

DOWNSTREAM

event-level runtime abstraction→hand-landmark model

可信度 95%

The abstraction is explicitly described as being 'between the hand-landmark model and the interaction task', meaning it consumes output from the hand-landmark model and transforms it for event-level confirmation.

REFERENCES

Transferable Latency Prediction for Fast LLM Screening on Heterogeneous Edge Devices→SmolLM2

可信度 95%

The paper states: 'the same SmolLM2 model family reaches 8.42 tokens/s on Orange Pi 5 Pro but 64.38 tokens/s on an RTX 3090-class GPU.'

REFERENCES

FlowEvo→HumanEval

可信度 95%

Experiments on benchmarks spanning interactive environments (ALFWorld) and code/math generation (HumanEval, GSM8K)...

REFERENCES

FlowEvo→ALFWorld

可信度 95%

On ALFWorld, FlowEvo achieves an 82.8% success rate, 23.6 percentage points above the strongest baseline...

UPSTREAM

Public-Space Gesture Interaction→hand-landmark model

可信度 95%

The paper states the system 'organize[s] recurring repair mechanisms into an event-level runtime abstraction between the hand-landmark model and the interaction task' and identifies the 'recognition-to-interaction gap' where frame-level recognition (hand-landmark model) fails to yield stable interaction events.

COOPERATES_WITH

FlowGuard→FlowVectors

可信度 92%

FlowGuard derives FlowVectors... that quantify cross-modal redundancy, synergy, and modality-specific dominance

COOPERATES_WITH

Transferable Latency Prediction for Fast LLM Screening on Heterogeneous Edge Devices→Orange Pi

可信度 90%

The framework is validated on Orange Pi 5 Pro, a device manufactured by Orange Pi.

COOPERATES_WITH

FlowEvo→DEFENSE-SEU

可信度 90%

The code is public at https://github.com/DEFENSE-SEU/FlowEvo.

UPSTREAM

FlowGuard→Multimodal large language models

可信度 90%

FlowGuard is a framework designed to secure multimodal large language models

COOPERATES_WITH

event-level runtime abstraction→Public-Space Gesture Interaction

可信度 90%

The abstraction is a core contribution 'aimed at bridging the recognition-to-interaction gap' in Public-Space Gesture Interaction; it is not a standalone system but an integrated component enabling the system’s event-level confirmation.

COOPERATES_WITH

Transferable Latency Prediction for Fast LLM Screening on Heterogeneous Edge Devices→Google

可信度 90%

The framework is evaluated on Pixel 8 and Pixel 8 Pro devices, which are Google products; the evaluation implies use of and alignment with Google's hardware ecosystem.

UPSTREAM

arXiv→FlowGuard

可信度 85%

The FlowGuard paper is published on arXiv (arXiv:2607.21600v1)