With the success of large language models (LLMs), integrating the vision model into LLMs to build vision-language foundation models has gained much more interest recently. However, existing LLM-based large multimodal models (e.g., Video-LLaMA, VideoChat) can only take in a limited number of frames for short video understanding. In this study, we mainly focus on designing an efficient and effective model for long-term video understanding. Instead of trying to process more frames simultaneously like most existing work, we propose to process videos in an online manner and store past video information in a memory bank. This allows our model to reference historical video content for long-term analysis without exceeding LLMs' context length constraints or GPU memory limits. Our memory bank can be seamlessly integrated into current multimodal LLMs in an off-the-shelf manner. We conduct extensive experiments on various video understanding tasks, such as long-video understanding, video question answering, and video captioning, and our model can achieve state-of-the-art performances across multiple datasets.

We introduce Patch Aligned Contrastive Learning (PACL), a modified compatibility function for CLIP's contrastive loss, intending to train an alignment between the patch tokens of the vision encoder and the CLS token of the text encoder. With such an alignment, a model can identify regions of an image corresponding to a given text input, and therefore transfer seamlessly to the task of open vocabulary semantic segmentation without requiring any segmentation annotations during training. Using pre-trained CLIP encoders with PACL, we are able to set the state-of-the-art on the task of open vocabulary zero-shot segmentation on 4 different segmentation benchmarks: Pascal VOC, Pascal Context, COCO Stuff and ADE20K. Furthermore, we show that PACL is also applicable to image-level predictions and when used with a CLIP backbone, provides a general improvement in zero-shot classification accuracy compared to CLIP, across a suite of 12 image classification datasets.

The ability to detect anomaly has long been recognized as an inherent human ability, yet to date, practical AI solutions to mimic such capability have been lacking. This lack of progress can be attributed to several factors. To begin with, the distribution of ``abnormalities'' is intractable. Anything outside of a given normal population is by definition an anomaly. This explains why a large volume of work in this area has been dedicated to modeling the normal distribution of a given task followed by detecting deviations from it. This direction is however unsatisfying as it would require modeling the normal distribution of every task that comes along, which includes tedious data collection. In this paper, we report our work aiming to handle these issues. To deal with the intractability of abnormal distribution, we leverage Energy Based Model (EBM). EBMs learn to associates low energies to correct values and higher energies to incorrect values. At its core, the EBM employs Langevin Dynamics (LD) in generating these incorrect samples based on an iterative optimization procedure, alleviating the intractable problem of modeling the world of anomalies. Then, in order to avoid training an anomaly detector for every task, we utilize an adaptive sparse coding layer. Our intention is to design a plug and play feature that can be used to quickly update what is normal during inference time. Lastly, to avoid tedious data collection, this mentioned update of the sparse coding layer needs to be achievable with just a few shots. Here, we employ a meta learning scheme that simulates such a few shot setting during training. We support our findings with strong empirical evidence.

A single chip radio transceiver for the DECT (Digital Enhanced Cordless Telephone) system has been implemented using Si 0.6 /spl mu/m BiCMOS technology. The device implements Rx down-converting mixers, a limiting IF strip, second down-converting image rejecting mixer, PLL demodulator with S-field sample and hold, bit slicer, synthesiser, UHF VCO, frequency doubler and transmit buffer. Synthesiser programming and the chip power management is achieved via the programming of an integrated 3 wire serial port. The chip also includes two separate supply voltage regulators, one of which is dedicated for use by the single UHF VCO.

This paper presents the generalised maximum signal-to-interference-plus-noise ratio (MSINR) array processor, which is the extension of MSINR from the single to the multiple desired-signal case. The problem of optimal multiple desired signal processing is of particular importance in a multipath fading environment. Therefore, the performance of the MSINR processor is evaluated for personal communication systems (PCS) in a Rayleigh fading channel with interference. Strong interference is expected in the PCS channel, due to the overlay in the spectrum allocated to PCS. It is shown that the generalised MSINR processor outperforms other methods the minimum variance distortionless response (MVDR) and the minimum mean square error (MMSE) interference cancellers, yielding the best bit error rate (BER).

Methodologies based on an elaboration of the knowledge acquisition and design structuring (KADS) philosophy were developed and a suite of tools based on this methodology was implemented. The tools implement object oriented support at the domain-layer, Bayesian reasoning combined with a Bayesian compatible version of "fuzzy sets" to support the inference-layer, and mechanisms for knowledge level task control to support both the task-level and strategic-level. These tools were built for the task of computerizing medical practice guidelines. The tools were successfully applied to two practice guidelines, one selected by symptom (febrile neutropenia) and the other by drug class (CSF).