Search arXivSearch

arXiv subjects

Tao Ling

Publications and source records attributed to Tao Ling.

3 recordsLinked to original sources

FedCoT-VQA: A Federated Learning and Unlearning Framework for Chain-of-Thought Planners in VideoQA

Chain-of-Thought (CoT) planners have emerged as an effective design for VideoQA, where a lightweight planner first generates intermediate reasoning steps to guide temporal evidence selection before answer prediction. This modularity makes CoT-based VideoQA attractive for federated learning, since only the planner side needs collaborative adaptation while the heavy vision-language backbone can remain fixed. However, in decentralized settings, the planner must not only be trained efficiently across heterogeneous clients but also support later client deletion requests. This is challenging because deleted-client influence is reflected both in model parameters and the planner's reasoning-trace behavior. We present FedCoT-VQA, a federated learning and unlearning framework for CoT planners in VideoQA. FedCoT-VQA consists of three modules: planner-side partitioning (PSP), which exposes a compact shared-residual adaptation space for efficient federated training; server-side aggregation (SSA), which aggregates planner-side updates while maintaining a deletion-ready contribution log; and a residual unlearning module (RUM), which approximates the retained-only counterfactual planner through retained-client replay and selective residual correction, without full retraining. We evaluate FedCoT-VQA in terms of federated training utility, federated unlearning utility, forgetting quality, and efficiency. Results show that compared to current federated approaches, FedCoT-VQA preserves strong federated training utility, improving grounding quality by up to 4.45%. After unlearning, it retains high accuracy and achieves a counterfactual gap of only 7.38%.

cs.CR

UV imprinting and aligned ink-jet printing for multi-layer patterning of electro-optic polymer modulators

The present work demonstrates an electro-optic polymer based Mach-Zehnder (MZ) modulator fabricated utilizing advanced ultraviolet (UV) imprinting and aligned ink-jet printing technologies for patterning and layer deposition. The bottom electrode layer is designed and directly ink-jet printed on the substrate to form the patterned layer. The waveguide structure is formed into a bottom cladding polymer using a transparent flexible mold based UV imprinting method. All other layers can be ink-jet printed. The top electrode is aligned and printed over the Mach-Zehnder arm. The modulator demonstrates a V-pi of 8V at 3kHz. This technology shows great potential in minimizing the fabrication complexity and roll-to-roll compatibility for manufacturing low cost, light-weight, and conformal modulators at high throughput.

physics.optics

Sensitivity enhancement in optical micro-tube resonator sensors via mode coupling

A liquid filled, low refractive index material inner-coated silica micro-tube has been proposed and studied as a coupled micro-resonator sensor to greatly enhance biochemical sensor sensitivity. Its unique coupling phenomenon has been analyzed and utilized to boost the refractive index sensitivity of the device to 967nm/RIU. Through optimization of the coupling strength between the two microresonators, further improvement in refractive index sensitivity up to 1100nm/RIU has been achieved. This mode coupling strategy allows us to design robust, thick-walled micro-tube sensors with ultra-high sensitivity which is useful in practical biochemical sensing applications.

physics.optics