Recent advancements in speech synthesis have leveraged GAN-based networks like HiFi-GAN and BigVGAN to produce high-fidelity waveforms from mel-spectrograms. However, these networks are computationally expensive and parameter-heavy. iSTFTNet addresses these limitations by integrating inverse short-time Fourier transform (iSTFT) into the network, achieving both speed and parameter efficiency. In this paper, we introduce an extension to iSTFTNet, termed HiFTNet, which incorporates a harmonic-plus-noise source filter in the time-frequency domain that uses a sinusoidal source from the fundamental frequency (F0) inferred via a pre-trained F0 estimation network for fast inference speed. Subjective evaluations on LJSpeech show that our model significantly outperforms both iSTFTNet and HiFi-GAN, achieving ground-truth-level performance. HiFTNet also outperforms BigVGAN-base on LibriTTS for unseen speakers and achieves comparable performance to BigVGAN while being four times faster with only $1/6$ of the parameters. Our work sets a new benchmark for efficient, high-quality neural vocoding, paving the way for real-time applications that demand high quality speech synthesis.

Despite the accelerated adoption of digital in recent years, some organisations are still trying to unlock true value from their transformation efforts to date – however as budgets compress and global economic uncertainty grows, many analysts are suggesting growth in the IT & Digital spend is essential for maintaining competitive advantage and staying operationally efficient. True digital transformation can only be seen when every function of an organisation comes together, providing the essential cohesion to innovate effectively and generate true transformation value. Whatever position a company is in, one thing that is clear is that the current generation of marketing and digital leaders is delivering cutting edge digital transformation programmes means continuing to break down organisational siloes wherever possible. This theme has been emphasised since everyone emerged from the pandemic. James McGough, Founder of Europe's biggest technology EXPO – DTX, believes crucial to this is the need for professionals to learn more about their own organisations: ''We have a lot of key discoveries in this area - 71% of our recent C level visitors have told us they no longer wanted to work in what I call'departmental isolation' and were actively seeking to work with others across their organisation to help realise the digital industry's massive potential." While leading companies can access a new world of possibilities by shifting to digital collaboration in all areas of their company, with technology stacks continuing to change rapidly, many marketing leaders are finding more obstacles to navigate than successes. Alex Vail, Chief Marketing Officer, R2 Factory at Rolls Royce feels more collaboration between all teams in order to share ideas, technology and data in new safe and secure ways is the answer to delivering real change: "Digital Transformation is hard.

With Ada Lovelace Day the inspiration for this special episode, we actually go on a journey beyond into'the how' of embedding diversity, equity and inclusion by design, and in these ever evolving times, how to better embrace change too. To do so, I am joined by the brilliant Isabelle Guis, Chief Marketing Officer at Commvault, who has a truly fascinating personal journey to share - I am sure it will inspire many! We cover the holistic skills that matter most today and into tomorrow including STEM and STEAM, growing through discomfort or'feel the fear and do it anyway', mottos for professional and personal life, the power of mentoring and sponsorship and so much more, packed with real life examples. And beyond this still, we explore the power of protection for data which alongside trust, must be the biggest currency of our times sharing security examples from Commvault and looking ahead to their Connections 2022 event on November 2nd and 3rd. This is a superb technology, learning and networking opportunity, freely available to all, that can help everyone to innovate with confidence.

Cloud can do much more than reduce a retailer's cost of computing and data storage. In fact, it can offer new experiences to both employees and customers which are hard to imagine without being on cloud. While this may sound intuitive to digital natives, traditional organisations need to wear a different lens to understand the power of cloud. Digital solutions can enable intelligent retail that better understands customers, empowers employees to deliver breakout customer experiences, builds robust supply chains and contributes to reimagining the customer experience. A good starting point is prioritising desired business capabilities.

Over the past two years, there's been an acceleration of the digital landscape in large part due to COVID-19. As a result, digital transformation has become front of mind for many businesses. Then it's important to understand that the key to success is linking commercial impact to data. This process includes connecting data across systems to generate a better user experience and then integrating that data to enable the business to measure and analyse effectiveness at scale. Done correctly, it provides visibility and allows data driven decision making at a velocity that generates business growth but also competitive advantages.

The problem of class imbalanced data lies in that the generalization performance of the classifier is deteriorated due to the lack of data of the minority classes. In this paper, we propose a novel minority over-sampling method to augment diversified minority samples by leveraging the rich context of the majority classes as background images. To diversify the minority samples, our key idea is to paste a foreground patch from a minority class to a background image from a majority class having affluent contexts. Our method is simple and can be easily combined with the existing long-tailed recognition methods. We empirically prove the effectiveness of the proposed oversampling method through extensive experiments and ablation studies. Without any architectural changes or complex algorithms, our method achieves state-of-the-art performance on various long-tailed classification benchmarks. Our code will be publicly available at link.

In the post-Covid era, organizations are recalibrating their marketing strategies to make better use of data and analytics to stay ahead of the competition. Increasingly, it is the Chief Marketing Officer (CMO) who has to take the lead to drive digital adoption and become the organization's digital evangelist. Data is everywhere and the CMO is increasingly expected to use multiple sources of data analysis and marketing intelligence for growth and revenue. Customer buyer journeys changed significantly during the pandemic: Buyers prefer to undertake a self-education journey, learning about the product rather than engaging with a salesperson right from the start. This puts the onus on marketers to provide prospective customers with the right content at the right time via the right channel.

A gossip protocol is a procedure for sharing secrets in a network. The basic action in a gossip protocol is a telephone call wherein the calling agents exchange all the secrets they know. An agent who knows all secrets is an expert. The usual termination condition is that all agents are experts. Instead, we explore protocols wherein the termination condition is that all agents know that all agents are experts. We call such agents super experts. Additionally, we model that agents who are super experts do not make and do not answer calls. Such agents are called engaged agents. We also model that such gossip protocols are common knowledge among the agents. We investigate conditions under which protocols terminate, both in the synchronous case, where there is a global clock, and in the asynchronous case, where there is not. We show that a commonly known protocol with engaged agents may terminate faster than the same protocol without engaged agents.

For the Chief Marketing Officer, the brand response to COVID-19 has been a case study in authenticity, focused on remaining true to the personality, spirit, character of their brand while keeping an eye on the real-world needs, fears, questions and concerns of their customers. Brands that have maintained open, bidirectional lines of communication and connection with customers are being repaid in loyalty and the promise of future profits, albeit delayed by the realities of a pandemic. We have seen the best of us rise to the occasion: Chiobani deploying food and resources to address food insecurity in a time of total insecurity, LVMH shifting from luxury parfum purveyors to churning out much needed hand sanitizer, and B2B brands like Zoho, Microsoft and SAP opening up resources to help aggregate data, empower the newly remote workforce and ensure that the voice of the customer (both internal and external) is still flowing into the heart of every organization. Sadly, we have also seen the worst: price gauging on cleansers, opportunistic advertising pushing fear, conspiracy and false hope cures, phishing scams preying on fear and hope, lazy automation inviting savings for in-store visits. Despite all the ups and downs, CMOs remain focused on mastering the 3R's of modern marketing, namely driving revenue, relationships and reputation.

Managing Marketing is a weekly podcast hosted by TrinityP3. Each one is a conversation with a marketing thought-leader, professional, practitioner or expert on the issues and topics of interest to marketers and business leaders everywhere. In this special series, TrinityP3's Anton Buchner, discusses the rise of Artificial Intelligence and the impact it is having on marketing. Jay Henderson is the Senior Vice President of Product Management at Acoustic (formerly Watson Customer Engagement – purchased from IBM by Centerbridge Partners, and rebranded in 2019 as Acoustic). He talks about how machine learning algorithms should be seen as working together with marketers. Offering options and solutions for marketers to assess and consider, rather than being seen as a distrustful'black box' of solutions running rampant by themselves. Welcome to Managing Marketing, a weekly podcast where we sit down and talk with marketing thought leaders and experts on the issues and topics of interest to marketers and business leaders everywhere. To discuss this I'm sitting down today with Jay Henderson. Jay is the senior vice-president of product management for Acoustic. Thanks, I'm really excited to be here. You've just flown in so you've got over your jetlag? I got here a couple of days ago. We're here today in Sydney to launch the Acoustic brand and the company into the Australian market.